Skip to main content
Internet Archive's 25th Anniversary Logo

Full text of "Pacific studies"

See other formats


a journal devoted to the study of the Pacific 
its islands and adjacent countries 

MARCH 1987 



Art History 






Political Science 


Published by 


(Brigham Young University — Hawaii Campus) 


Fergus Clunie 
Paul Alan Cox 
Roger Green 
Renee Heyum 
Francis X. Hezel, S.J. 
Rubellite Johnson 
Adrienne Kaeppler 
Robert Kiste 
Robert Langdon 
Ioane LeMamea 

Stephen Levine 
Katharine Luomala 
Barrie Macdonald 
Cluny Macpherson 
Leonard Mason 
Malama Meleisea 
Norman Meller 
Richard M. Moyle 
Colin Newbury 
Douglas Oliver 
Margaret Orbell 
Nancy Pollock 
Karl Rensch 
Bradd Shore 
Yosihiko Sinoto 
William Tagupa 
Francisco Orrego Vicuna 
Tuaopepe Felix Wendt 
Edward Wolfers 

Fiji Museum 

Brigham Young University 

University of Auckland 

University of Hawaii 

Micronesian Seminar 

University of Hawaii 

Smithsonian Institution 

University of Hawaii 

Australian National University 

Pacific Islanders' 

Educational Research Center, N.Z. 

Victoria University 

University of Hawaii 

Massey University 

University of Auckland 

University of Hawaii 

University of the South Pacific 

University of Hawaii 

University of Auckland 

Oxford University 

University of Hawaii 

Canterbury University 

Victoria University 

Australian National University 

Emory University 

Bishop Museum 

Office of Hawaiian Affairs 

Universidad de Chile 

University of the South Pacific 

Macquarie University 



Associate Editor Associate Editor 


Book Review Editor 

Editorial Policy 

Pacific Studies is published three times each year by The Institute for 
Polynesian Studies, Brigham Young University — Hawaii Campus, Laie, 
Hawaii, 96762, but responsibility for opinions expressed in the articles 
rests with the authors alone. 

Subscription rate is U.S. $20.00. Accounts payable to The Institute for 
Polynesian Studies should be sent to the editor. Articles submitted to the 
editor should be the original typewritten copy, double spaced. Authors 
may write to the editor for a style sheet. Books for review should also be 
sent to the editor. 

The Institute for Polynesian Studies is an organization funded by 
Brigham Young University — Hawaii Campus. The Institute assists the 
University in meeting its cultural and educational goals by undertaking 
a program of teaching, research, and publication. The Institute cooper- 
ates with other scholarly and research institutions in achieving their 
objectives. It publishes monographs, produces films, underwrites 
research, and sponsors conferences on the Pacific islands. Further infor- 
mation on the activities of the Institute may be obtained by writing to 
its director, Jerry K. Loveland, at the above address. 

©1987 Brigham Young University — Hawaii Campus. All Rights 
Reserved. Printed in the United States of America. 

ISSN 0275-3596 

Digitized by the Internet Archive 

in 2012 with funding from 

Consortium of Church Libraries and Archives 

Volume 10 March 1987 Number 2 


Reality and Fantasy: The Foster Child in Hawaiian Myths 
and Customs 
Katharine Luomala 1 

The Atomization of Tongan Society 

Keith L. Morton 47 

The Lives and Times of Resident Traders in Tuvalu: 
An Exercise in History from Below 
Doug Munro 73 

Book Review Forum 

Valerio Valeri, Kingship and Sacrifice: Ritual and Society 
in Ancient Hawaii 

Review: John Charlot 107 

Response: Valerio Valeri 148 

Contributors 215 


Vol. 10, No. 2 March 1987 


Katharine Luomala 
University of Hawaii 

Hawaiians, like other Pacific islanders, narrate prose sagas that incor- 
porate motifs of the quest of a character, almost always male, to learn 
the identity of and then to locate his biological father. 1 As the father is 
either a great chief or a god, the youth expects to receive the privileges 
and rights that he believes are his birthright. Nine Hawaiian characters 
— eight male, one female (patterned after the male) — ask their female 
caretaker, "Who is my father?" Although the query is expressed in two 
tale- types also present elsewhere in the Pacific, no other archipelago has 
as many different semihistorical, mythical, and fictitious heroes who 
ask about their unknown father as the Hawaiian Islands. 2 

The principal tale-type as developed in Hawaii begins with a roving 
chief or sky god having a romantic encounter with a woman usually of 
lower rank than himself. Before going home he gives her tokens of his 
rank and a name for their anticipated son. She rears the child alone or, 
more commonly, with her brother or husband, who punishes the child 
for behaving as if he were a chief. The unhappy child, feeling himself a 
misfit in this environment, asks his mother: "Who is my father?" Eva- 
sive at first, she finally tells him the truth and advises him how to find 
his father. When he has located him he proves, often with difficulty at 
first, that he is indeed his son and entitled to his birthright. The saga 
often portrays him next as a usurping chief who displaces his elder half- 
brother or a god to gain great power. He may also marry his royal half- 

Pacific Studies, Vol. 10, No. 2— March 1987 

2 Pacific Studies, Vol. 10, No. 2— March 1987 

sister to have offspring of higher rank than himself. But it is the early 
events in the hero's career that are the major focus of interest here. 

The hero of the second tale- type is always named Laka. When his 
playmates make him aware that unlike them he has no father, he asks 
his mother, "Who is my father?" He learns that his father was murdered 
on another island while seeking a birth gift for him. Laka determines to 
build a canoe and go off to gain revenge and bring home his father's 
bones. After forest spirits restore a tree each time he has irreverently 
chopped it down for a canoe, his mother or grandmother tells him how 
to placate them. Once the spirits are appeased, they make his canoe and 
he leaves with supernatural companions to successfully avenge his 
father's death and bring home his bones. 3 Getting the bones is impor- 
tant for otherwise enemies might dishonor them by making them into 
fishhooks or other artifacts. 

Psychoanalysts, psychologists, social workers, and others have ob- 
served in both normal and neurotic children and adults, male and 
female, what they call the "foster-child fantasy," according to which an 
individual believes that he is not the child of the ordinary couple who 
are rearing him but is the offspring of renowned, albeit unknown, par- 
ents. The motif of a male foster child who is reared without knowing 
that his natural parents are great personages occurs in many Old World 
and Near Eastern myths that Freud and his followers, particularly Otto 
Rank, have interpreted in terms of their psychoanalytical theories. 
Because in the Hawaiian sagas only the father's identity is unknown to 
the child, it is of interest to analyze them in relation to the psychoana- 
lysts' theories — which are based on a different culture — and as a reflec- 
tion of traditional Hawaiian social customs with regard to parentage, 
rank, adoption, and fosterage. First, I will present a selection of rele- 
vant theories of Western psychoanalysts and psychologists based on 
Euroamerican and Near Eastern culture. 

Psychoanalytical Theories of the Foster-Child Fantasy 

Otto Rank, while recognizing the importance of cultural variability and 
diffusion in narratives about the quest for the unknown father, empha- 
sized that their ultimate origin is in the human psyche. 4 The myth- 
maker, he wrote, created the myth from retrograde infantile fantasies 
and credited the hero with his own personal infantile history. Rank dis- 
cerned a pattern, despite differences in expression, in accounts of the 
lives of fifteen mythic or semimythic Old World heroes. The noble par- 
ents have some difficulty with conception, such as prolonged barren- 

The Foster Child in Hawaiian Myths and Customs 3 

ness, continence, or taboos on intercourse. When the hero is born they 
have him exposed, frequently to a watery death, because before or dur- 
ing the pregnancy a prophetic dream or an oracle has cautioned against 
his birth, usually because of danger that will befall his father. Saved and 
reared by lowly foster parents or animals until grown, the youth then 
leaves and after unusual adventures finds his true parents. Although 
they may acknowledge him as their son he may nevertheless take 
revenge, especially against his father, for having abandoned him. He 
may later win honors and fame but his life may end in tragedy. Of the 
many variations of the pattern, the best known is the Oedipus myth, in 
which the hero unwittingly slays his father and marries his mother. 

The hero, according to Rank, represents Ego acting out in narrative 
form the Freudian "family romance" of real life. The young child at 
first idealizes and overvalues his parents as all-powerful and wonderful. 
However, as he grows more independent he becomes disillusioned by 
their ordinariness, real or imagined neglect, discrimination, or unsym- 
pathetic treatment. He then fancies, sometimes under the influence of 
romantic stories, that he is not their child but that of exalted persons. 
The latter represent the idealized parents he once thought his true par- 
ents were. As his sexual awareness develops, his strong sentimental 
attachment to his mother leads him to alter his fantasies so that he now 
wishes to eliminate competition for her affections. He therefore rebels 
against his father with the subconscious intent of displacing him. Never- 
theless, he usually conceals his forbidden erotic wishes by trying to free 
himself from parental authority, develop a mature, integrated personal- 
ity, and achieve his goals. As he grows older he consciously but ambiva- 
lently criticizes his fantasy while subconsciously clinging to it. Even as a 
normal adult he may revert to it in dreams. The male's Oedipus com- 
plex thus described has its counterpart in the female's Electra complex. 

Fantasies of foster parentage and noble lineage led some of Philip R. 
Lehrman's patients to delusions of grandeur of being a crown prince or 
princess deprived of a rightful heritage. 5 Among milder expressions of 
their fantasy, these individuals evinced great interest in their genealogy, 
were estranged from their family, denied their nationality, changed 
their name, joined secret societies, or either sought or bestowed titles 
and degrees. An individual's fantasy might be released by a national 
upheaval or a major shift of residence. An immigrant, for instance, 
might revise his personal history in order to claim that he had high 
social position and estates in his homeland. On the other hand, the fan- 
tasies of a normal individual generally became the basis of constructive 
plans and actions, not of deception as an imposter or poseur. 

4 Pacific Studies, Vol. 10, No. 2— March 1987 

To test the theory that all or most children have the foster-child fan- 
tasy during the development of the "family romance" and are influ- 
enced by it, Edmund S. Conklin in 1920 issued a questionnaire to 904 
male and female students of late high school and early college age. 6 In 
the sample 28.5 percent had voluntary immediate recall of the fantasy, 
mainly in an incipient, unclear stage as a belief, daydream, or casual 
thought. More than half had it between the ages of eight to twelve. 
Slightly more girls than boys recalled it. The fantasy lasted either a brief 
time or as long as two years, with an occasional respondent still clinging 
to it. Almost as many who at first believed themselves of inferior as 
opposed to superior origin revised that fantasy to include greatness as 
they became more independent and lost the feeling of helplessness that 
had made them believe themselves orphans or foundlings. 

Some respondents believed that their "real" parents were wealthy, 
noble, famous, or royal, or were supernatural beings. There were also 
those who thought that they were of status similar to that of the 
assumed foster parents. Many gave additional reasons for their fantasy 
that were not on Rank's list of real or imagined parental mistreatment 
and neglect, lack of affection, and suggestions from stories. These addi- 
tional reasons included philosophizing (especially by boys) and knowing 
or hearing of actual cases of adoption and fosterage. Other reasons were 
dissatisfaction with economic limitations (usually by those with the fan- 
tasy of highborn parentage), absence of mental and physical family 
resemblance (especially by girls), encounters with famous people, and 
peculiar family circumstances (stepparent, family quarrels, marital 
infidelity, and parents' prolonged absences). Some individuals, having 
conceived the idea, sought proof, but others became depressed, ran 
away, and engaged in alienating behavior. Still others tried to behave 
better through gratitude to their assumed foster parents. 

Since the psychoanalysts and psychologists have formulated and illus- 
trated the foster-child fantasy almost entirely on the basis of Euroameri- 
can and Near Eastern society and literature, it is of interest to see how 
both fantasy and reality are expressed in the narratives, beliefs, and cus- 
toms of a different society, namely the Hawaiian, in which paternity 
was frequently actually in doubt. 

Nine Hawaiian Father-Seekers 

Of the nine Hawaiian characters asking "Who is my father?" because 
they do not believe that their male caretaker is their biological father, 
only the demigods Maui and Laka are known outside the Hawaiian 

The Foster Child in Hawaiian Myths and Customs 5 

Islands, and only Laka's father is dead. 7 The Hawaiian Maui is the only 
Hawaiian father-seeker born in nonhuman form. His mother, Hina-of- 
the-fire, bore him as an egg, which hatched into a crowing cock that 
then became a male child. 

Three heroes who actually lived in the Hawaiian Islands around the 
fifteenth or sixteenth century a.d. are 'Umi-a-Liloa, son of Liloa, the 
sacred paramount chief of northwestern Hawaii, and a very low-rank- 
ing chiefess; 'Umi's future brother-in-law Kiha-a-Pi'ilani, son of Pi'i- 
lani, the ruling chief of eastern Maui, and his highborn queen; and 
Paka'a, son of a commoner named Kuanu'uanu, King Keawe-a-'Umi's 
chief steward, navigator, and councilor. Kuanu'uanu named his expect- 
ed son by a minor chief's daughter for Keawe's scaly and wrinkled 
(paka'a) skin caused by his excessive indulgence in kava, a chief's privi- 
lege. Both Kiha and 'Umi became great kings and conquerors while 
both Paka'a and his son served Keawe, as did Kuanu'uanu. In genealo- 
gies Maui, 'Umi, and Kiha are listed among the ancestors of Kings 
Kamehameha I and Kalakaua. There appears to be no genealogy for 
Paka'a, perhaps because his low rank did not permit him to keep one, 
which only chiefs could do. A high chief might give a valuable com- 
moner a nominal title of chief as Keawe did to Paka'a and his father, but 
such titles were outside the category of inherited rank as a chief. 

Three heroes and a heroine are completely fictitious and no one 
claims descent from them in royal genealogies. Unlike the matter-of- 
fact narrative style of the sagas about 'Umi, Kiha, and Paka'a, the sagas 
about the four fictitious characters are florid in style, the incidents fan- 
tastic, and supernatural characters much in evidence. Each of the four 
has a highborn father and a mother who is usually a commoner. One 
hero's mother, however, is exceptional in being the supernatural queen 
of a floating island where she has a romance with an earthly ruling 
chief. Their son's name, Na-ku'e-maka-pau-i-ke-ahi (Eyebrows burnt 
off in the fire) , commemorates the chief's accident while trying to teach 
the queen to make fire and cook food, which were not known on her 
island. Like several other romantic and precocious heroes, Naku'emaka 
is born three years after his father's departure, walks on the second day, 
talks on the third, and plays darts on the sixth day with big boys who 
taunt him for being fatherless. 

Another fictitious hero is Kalani-manui'a (High Chief Manui'a), 
whose name reflects his father's pride in his own rank and that of his 
anticipated son despite the mother being a commoner. How Nl'au-e- 
po'o, a cloudland king's son, got his name and what it means is uncer- 
tain. 8 If it were Nl'au-'e-'e-po'o (Coconut leaf midrib climbing to the 

6 Pacific Studies, Vol. 10, No. 2— March 1987 

summit) it would refer to the boy's journey to his sky father on a stretch- 
ing coconut tree (a familiar narrative device). 9 Actually Nl'au just sat on 
the tree as it rose, he did not climb it, but he is not the only hero or hero- 
ine to do that in a story. The fourth fictitious character is a female 
named Lau-kia-manu-i-Kahiki (Bird trapping leaf in Kahiki), her 
cloudland father's name for his anticipated daughter by a commoner. 

I have located thirteen published narratives in which each of the nine 
father-seekers asks directly or indirectly about the father's identity. 
Three versions with the query relate to 'Umi; two each to Laka and 
Paka'a; and one each to Maui, Kiha, Kalanimanui'a, Nl'au, Laukia- 
manu, andNaku'emaka. 10 

Although only three narrators have 'Umi ask about his father, none of 
the numerous other references to him express any doubt of his being the 
son of Liloa's misalliance with a low-ranking woman. Perhaps the other 
narrators regarded 'Umi's query as superfluous since, like the three 
storytellers who include it, they have 'Umi's mother protest to her hus- 
band that the boy he is beating is not his son but King Liloa's. The 
mother's protest naming the natural father is sufficient for 'Umi to seek 
out King Liloa. 

All narrators agree that Kiha, the younger son of the ruling chief of 
eastern Maui and his highborn wife of an important Oahu family, was 
reared by his royal mother and her kin at the court of Queen Kukani- 
loko on Oahu. However, only one narrator describes Kiha's boyhood 
and states that when the boy's maternal uncle scolded him for wasting 
food he asked his mother about his absent father. It seems unlikely that 
the son of two such prominent persons would never have heard at the 
Oahu court that his father was Chief Pi'ilani of Maui. This narrator, 
like the three who have 'Umi ask about his father's identity, is trying to 
adapt another famous legendary chief to the popular hero pattern. He 
has used stock incidents to fill a gap in knowledge about the boy Kiha's 
relationships with his mother and other caretakers. 

Obscurities in those variants with the child's query can be illuminat- 
ed by those without it. The demigod Maui's query, for instance, occurs 
only in an ambiguous name chant that is a cryptic biography of his life 
from birth to death, listing his struggles to gain power and usurp the 
privileges of gods. The chant is part of the Kumulipo, a genealogical 
creation chant compiled in its present form perhaps around a.d. 1700. 
Hina is puzzled because although she has slept with neither a fowl nor a 
man she has given birth to an egg that hatched into a cock and became a 
male child, called Maui-of-the-malo. A malo is a frequent symbol for 
sexual intercourse, and the cock, of a usurping chief. After defying his 

The Foster Child in Hawaiian Myths and Customs 7 

maternal uncles and the gods Kane and Kanaloa, Maui asks his mother 
who his father is. She denies that he has a father, saying that the loin- 
cloth of Akalana (Kalana) is his father. The fact that she sends him to his 
makua kane, male parent, for a fishhook called Manai-a-ka-lani (Nee- 
dle of the sky) hints that his father is a sky god. The difficulty in this 
context is that makua kane may designate the father, uncle, or male 
cousin of either the mother's or the father's generation. 

However, a fragmentary prose narrative that lacks the query identi- 
fies the father with the sky, for he is Makali'i (Pleiades), to whom Maui 
travels on a stretching coconut tree. The latter is his transformed mater- 
nal uncle who earlier took the shape of a canoe to transport Hina to her 
dream lover, Makali'i, then on Kauai. The two had eight children, 
including Maui, before they separated. 11 

A different myth without the query about the father also throws light 
on the name chant. In one of its two variants, Kane and Kanaloa magi- 
cally impregnate Hina when, while bathing, she puts on the malo of 
Chief Kalana-mahiki of Hilo. In this fragment the egg Hina bears 
becomes Maui, whom she sends, when he has grown up, to live with his 
father, Kalana-mahiki, and his half-brothers. The two tokens — the 
chief's malo and his staff — that she now gives him were not mentioned 
earlier. 12 There the story ends. The other variant, while fuller, lacks the 
tokens, the egg, and the malo owner's name. The owner, however, is 
obviously a chief because the malo is red, the color of chiefs. And when 
Hina tells her husband, Akalana, about the red malo he says, "We shall 
have a lord." 13 

These two variants suggest that Akalana (Kalana) is Hina's nominal 
husband but not Maui's real father. Narratives about Maui, like those 
about 'Umi and Kiha, have assimilated him to the hero pattern to give 
him a more distinguished father than his mother's less impressive hus- 
band. The process in which Maui asks who his father is was already well 
under way by a. d. 1700. 

Two heroes whose careers resemble those of the nine characters and 
have been based on the traditional hero pattern have not been included 
in this study. This is because they do not need to ask "Who is my 
father?" 14 They have always known his identity. Each, taking his 
father's tokens with him, leaves his mother and stepfather or foster 
father to join his natural father in the sky. The father of one of the two 
heroes asks the mother to use the name Ke-au-nini-'ula-o-ka-lani (The 
restless red current of heaven) if their anticipated child is a boy. 
Keaunini does not ask his mother who his father is but only how to 
reach him. When he meets his father they fight and Keaunini kills him 

8 Pacific Studies, Vol. 10, No. 2— March 1987 

because the father did not give him a chance to display the tokens and 
prove his relationship. The other hero, Na-maka-o-ka-pao'o (The eyes 
of the goby), presumably was also named by his father but the narrator 
does not say so. The storyteller, obviously beginning to create a new 
saga from old materials, neither integrated nor completed his story. 
How the boy learned his father's name and whereabouts is not divulged 
nor even how he reached his cloudland father. 

The following outline gives the principal elements of those parts of 
the sagas that concern the parentage, childhood, and meeting with the 
father of each of the nine characters. Excluded are both the father's 
numerous adventures that led to his meeting the mother and the hero's 
later career. Laka, son of Wahieloa and Hina, and Kiha, son of Pi'ilani 
and La'ielohelohe, are absent from Part I of the outline because unlike 
the other characters they were not conceived at a clandestine meeting; 
they were born to parents who were already married to each other. 

Part I. The Parents' Meeting and Separation 
A roving chief ("king") of Hawaii, Oahu, or a cloudland has an affair 
with a woman at her bathing place. When she becomes pregnant he 
leaves her with a name for the expected child and tokens of rank by 
which it can later prove its paternity. The father may also instruct the 
mother about how the child should present itself to him in order to be 
recognized. On returning to his domain he may prepare amenities for 
the child's expected arrival. (The child's name is given below in paren- 

The Couple 

• Father. An earthly chief or king: Liloa ('Umi); Kaewaeoho (Naku'e- 
maka); Ku (Kalanimanui'a); Akalana or Kalana-mahiki (Maui, 
variants). A commoner: Kuanu'uanu (Paka'a). A supernatural sky 
god or king: Kane and Kanaloa or Makali'i (Maui, variants); 
Kualaka'i (Nl'au); Maki'ioeoe (Laukiamanu). 

•Mother. A commoner: Kauno'a (Kalanimanui'a); Hina (Nl'au); 
Hina (Laukiamanu). A minor chiefess: La'amaomao (Paka'a); Aka- 
hikuleana ('Umi). A supernatural being: Hina-of-the-fire (Maui); 
Kaanaelike (Naku'emaka). 15 

Father's Instructions 

• Child's Name. Formula: "If a boy, name it (father's choice), if a 
girl, name it for your side." Exception: "If a girl, name it Laukia- 
manuikahiki, if a boy, Maki'ioeoe." 

• Orders for Child's Recognition. Child is to present tokens to its 
father. Exceptions: No tokens left for Paka'a, Naku'emaka, and 

The Foster Child in Hawaiian Myths and Customs 9 

Maui (except in one variant). Child must also arrive in a red canoe 
with red sails, red crew, and red gear (Nl'au, Laukiamanu). 

• Tokens. Malo, warclub, whale-tooth necklace ('Umi); malo, spear 
(Kalanimanui'a); red feather helmet, cape, canoe (Nl'au); feather 
cloak, whale-tooth necklace, bracelet (Laukiamanu); chief's malo 
and staff (Maui, variant). 

• Father's Preparations. Taboos his royal wife's daughter for expected 
son's wife (Nl'au); taboos bathing pool and other amenities for 
expected child (Nl'au, Laukiamanu). 

Part II. The Child's Upbringing 
The child, reared by the mother and her kin, feels a misfit. Its uncle or 
stepfather mistreats it, or if reared only by the mother it notices that 
other children have a father. The child may have more than one reason 
for asking the mother who its father is. Sometimes evasive at first, she 
identifies the father, produces the tokens (if any), and advises the child 
about the journey and how to behave toward its father. 


• Primary caretaker. Mother ('Umi, Kiha, Paka'a, Naku'emaka, 
Kalanimanui'a, Nl'au, Laukiamanu, Maui, and Laka in one vari- 
ant); grandmother (Laka, in another variant). 

• Secondary caretakers. Stepfather ('Umi, Laukiamanu, Kalanima- 
nui'a, Maui). Maternal kin: grandparents, ancestor (Nl'au); grand- 
mother, grandaunt (Laukiamanu); uncle ('Umi, Kiha, Paka'a); un- 
cles (Maui, one variant); mother or grandmother (Laka, variants). 

Child's Queries to Mother about Father 

• Age. Six days old (Naku'emaka); four years old (Paka'a); "not yet 
grown up" (Laka); pre- adolescent or adolescent (implied for 
others) . 

• Reasons for Queries. Observes playmates have a father (Laka, 
Nl'au, implied for Naku'emaka); reflects, wonders, reasons about 
father ('Umi, Paka'a, Maui); wants more power after defeating 
uncles and gods (Maui); hears mother tell stepfather ('Umi, Kalani- 
manui'a); feels real father would be kinder (Kiha, Laukiamanu, 
implied for 'Umi); unhappy because male caretaker punishes him 
for generosity with food to playmates ('Umi, Kalanimanui'a, Lau- 
kiamanu), eating before him ('Umi), eating too much ('Umi, Kiha), 
destroying plants (Kiha), all characteristics of a chief. 

• Number of Queries before Learning Truth. One ('Umi, Kiha); two 
(Naku'emaka, Maui, Laka); asks for ten days (Kalanimanui'a); asks 
persistently (Paka'a, Nl'au, Laukiamanu). 

10 Pacific Studies, Vol. 10, No. 2— March 1987 

• Form of Query. 'Umi: "Have I not a different father?" Or, "Have I 
no other father but this one?" "Is he my only makua?" Or, "Is this 
indeed my father?" Kiha: "Where is my father? This is not my 
father. He is a man who gets angry with me." Paka'a, after rejecting 
statement that his maternal uncle, a small boy, is his father: "Who 
is my father and where is he?" Or, he "began to wonder where his 
father was." Kalanimanui'a: "Who then is my father?" Naku'e- 
maka, after playing with boys who have a father and saying that he 
knew he had a father because he knew the reason for his name, 
asks, "Where is my father?" Nl'au "asked where his own father 
might be." Maui "reflected, asked who was his father." Laka: "How 
is it that I have no father while other children have one?" Or, his 
long-lost father "is asked about by Laka." Laukiamanu "began 
questioning who her own father was until the mother could stand it 
no longer." 

Mother's Initial Reply if Evasive. To Kalanimanui'a: "You have 
no father, this (her husband) is your father." To Nl'au: "Alas! He is 
dead, only we two are left." To Laukiamanu: "The cliff is your 
father" (cliff denies this); "the bamboo is your father" (bamboo 
denies this, names the father); the mother confirms this. To Naku'e- 
maka: "You have no father." To Paka'a: "Mailou (her young 
brother) is your father." To Maui, "You have no father." To Laka: 
"Ask your grandmother." 

Mother's Aid in Finding Father 

• Gives Material Objects. Father's tokens ('Umi, Kalanimanui'a, 
Maui, variant; Nl'au forgets to take them); calabash of winds and 
bundle with white malo, fine grass cape, fan (Paka'a); red canoe 
and escort on her rolling island (Naku'emaka); canoe and food 

• Gives Other Help. Sends her young brother 'Oma'okamau as atten- 
dant to carry warclub ('Umi); explains route, proper behavior to be 
acknowledged by father— namely, sit on his lap (permitted only to 
his own children), tell name, show tokens if any ('Umi, Kiha, 
Paka'a), but Kiha as younger son is to sit on father's left knee, take 
food and drink from his left hand, as the right side belongs to older 
son; sends child to her elders for help (Nl'au, Laukiamanu, Laka); 
advises how to placate forest spirits (Laka, variant) . 

• Mother's Helpful Elders. Her parents give direction-finding bow 
and arrow (Nl'au); shape-shifting ancestor or uncle takes child to 
sky (Nl'au, Maui in one variant); child's blind, banana-cooking 
grandmother and grandaunt provide stretching bamboo to sky 

The Foster Child in Hawaiian Myths and Customs 1 1 

(Laukiamanu); grandmother advises how to placate forest spirits to 
get canoe and four supernatural men become his crew (Laka, vari- 
ants) . 

• Mother's Lack of Help. Neither mentions nor gives tokens to daugh- 
ter (Laukiamanu). 

Part III. Journey and Arrival at Father's Court 
On the journey to an earthly or sky father, the child's adventures, if any, 
may be pleasant or hazardous. No child is immediately accepted by its 
father. It may even be killed and later magically restored to life when 
priests identify its spirit. Ceremonies and marriage may follow but may 
be marred by the mother's revenge against the father for initially reject- 
ing their son. There is no information on Maui's reception. 

Events Enroute. Mother recites protective spells as stretching tree 
raises frightened boy to sky (Nl'au); child wins games with boys in 
sky (Nl'au); acquires friend of same sex in sky (Nl'au, Laukiamanu); 
adopts one or more boys as sons and companions ('Umi); joins crew 
of sight-seeing king to reach Hawaii (Paka'a) . 
Arrival at Court 

• Father's Initial Rejection. Has guards seize or try to seize apparent 
taboo-breaker ('Umi, Paka'a, Laukiamanu, Naku'emaka); drops 
child from lap ('Umi), or tries but fails to drop child from lap 
(Paka'a); surprised at younger son sitting on his right knee and tak- 
ing food from his right hand (Kiha); has child killed (Kalanima- 
nui'a's body thrown in ocean; bodies of Nl'au and friend to become 
burnt sacrifices; Laukiamanu thrown in pigpen, to be killed and 
baked later); no information (Maui); encounters witch guarding his 
dead father's bones (Laka) . 

• Reasons. Child has no tokens or red canoe (refused to go in one); 
uses tabooed amenities (Nl'au, Laukiamanu); marries tabooed half- 
sister (Nl'au); gets no chance to show tokens or tell name (Kalani- 
manui'a); is unrecognized on lap ('Umi, Kiha, Paka'a, Naku'e- 

Father's Acknowledgment. Sees his necklace on child ('Umi); hears 
son's name (Kiha, Paka'a); recognizes son when he shifts to his right 
knee (Paka'a); priest identifies child (Naku'emaka); when priests 
say the slain child's spirit rises from sea each night, father has it 
netted (Nl'au, Kalanimanui'a); ancestor Niu-ola-hiki (Life extend- 
ing coconut tree), who raised child to sky, now takes eel form to 
have sea gods resuscitate Nl'au and turn friend into a red fish; 
priests partially restore and resuscitate slain child's body (Kalani- 

12 Pacific Studies, Vol. 10, No. 2— March 1987 

manui'a); grandaunt in owl form brings tokens, chants parentage 
(Laukiamanu); no information on means of identification (Maui). 
Some Subsequent Events 

• Ceremonies. Symbolic navel-cutting ceremony to confirm son's 
identity ('Umi); feast of celebration (Naku'emaka); faulty rites to 
appease Niuolahiki, a factor in father's death (Nl'au); beauty con- 
test, magically judged, selects completely restored son as winner 

• Revenge. Father orders guards killed and baked for not recognizing 
child (Naku'emaka, Laukiamanu); mother angry over father's ini- 
tial rejection kills him, he becomes a fish (Nl'au); mother would 
have killed father but son helps him with wife-identification contest 
(Naku'emaka); 16 Laka and his crew kill the guardian of his father's 
bones and take them home. 

• Trouble with Siblings. Elder half-brother angry at child usurping 
his privileges and rights; to escape death when brother becomes 
king, hero hides incognito in exile; priests recognize him, help him 
kill and depose half-brother and become king ('Umi, Kiha); broth- 
ers' broken taboo causes magically hooked islands to scatter (Maui). 

•Marriage. Hero weds half-sister ('Umi, Nl'au, Kalanimanui'a); 
weds half-brother (Laukiamanu). 

Psychological Factors in the Sagas 

Each Hawaiian male character's puzzlement about his true father 
recalls the foster-child fantasies of Conklin's American respondents and 
Freud's and Rank's theories, based on Old World and Near Eastern 
myths, that narrators project into each character's life their own child- 
hood fantasies based on the "family romance." Of Conklin's respon- 
dents, more girls than boys reported having the fantasy, but in tradi- 
tional narratives in Hawaii and other parts of the Pacific males are 
almost always the father-seekers. In Polynesian narratives, in fact, I 
have come across only three examples of a female quest for the unknown 
father; each example, whether from Hawaii, New Zealand, or the 
Tokelaus, is differently developed. 17 The saga about the Hawaiian girl, 
Laukiamanu, was, I suggest, adapted from those about male father- 
seekers, most specifically that about Nl'au; the Nl'au saga, in turn, may 
have been inspired by that about Kalanimanui'a. The three sagas con- 
stitute a unit with numerous detailed similarities beyond those they 
share with the other sagas. 

When the child asks the decisive question, "Who is my father?" that 

The Foster Child in Hawaiian Myths and Customs 13 

will change his life, he is, except for very precocious youngsters like 
Naku'emaka and Paka'a, at a pre-adolescent or adolescent stage like 
Conklin's respondents. The child's uncertainty about his natural father, 
some storytellers suggest, originated earlier and then crystallized. The 
same applies to Conklin's respondents except that they thought they had 
been reared by two adoptive parents, not by a surrogate father and a 
natural mother. 

Like Conklin's respondents, the Hawaiian characters had other rea- 
sons for asking the question besides their tendency to reflect and feel 
unhappy and alienated. Like the Old World and Near Eastern heroes, 
the Hawaiians were reared under humble or unusual circumstances, 
but the Hawaiian reasons usually developed from the child being 
punished by the surrogate father for chieflike behavior relating to food 
when, presumably, he was a commoner (except for Kiha). A chief 
always had plenty of food to eat but had the responsibility to distribute 
some to his dependents in return for their having produced it and hav- 
ing otherwise served him. The child ate too much, wasted food, de- 
stroyed plants, distributed family food generously to his playmates, or, 
feeling superior to his surrogate father, ate before he did. Chieflike 
qualities were also evident in his skill in surfing and other sports. Except 
for a child like Laka reared in a one-parent home, the father-seekers 
appear to have been well-liked by their peers, doubtless because of their 
generosity. In the one-parent home without a male authority the child's 
chieflike quality was exhibited by his success in games, which led his 
playmates to taunt him as fatherless. 

Most Hawaiian father-seekers had a substitute male parent who was 
either the mother's husband or one of her brothers. The substitute is a 
commoner or, like Kiha's uncle, of lower rank than the natural father. 
The image of the surrogate father is negative, and perhaps to further 
degrade him few narrators bother to give his name. He and the child, 
whether male or female, are mutually hostile, with food as the most fre- 
quent source of overt dispute. Psychologically, he displaces the real 
father as the object of the growing child's rebellion and disillusionment. 
Nonetheless, after assuming his rightful rank, the child rarely takes 
revenge for earlier mistreatment. After all, the man had mistakenly 
assumed (except in the case of Kiha and perhaps Maui) that the child 
was his and must be taught behavior appropriate to the humble way of 
life he would always lead. He and the child are not only hostile to each 
other but probably jealous, for there is a very close and loving bond 
between mother and son. The substitute father's role highlights the 
child's supernatural or chieflike qualities, provides a contrast with the 

14 Pacific Studies, Vol. 10, No. 2— March 1987 

natural father who represents the ideal, and motivates the son to ask if 
he has no other father but this man. The answer to the query propels 
him out into the world to achieve his destiny. 

Each father-seeker, male or female, is either the mother's only child 
or narrators have obscured the presence of half-brothers and half-sisters 
who would compete for the mother's affection. Maui, known to have 
brothers, is an exception., but the Kumulipo, after naming them, com- 
pletely ignores them. In prose narratives, however, they function 
importantly as Maui's foils and marplots. Sibling rivalry is most devel- 
oped as a theme after the son has been accepted at his natural father's 
court. He must then contend with mistrust, jealousy, and physical dan- 
ger not only from his half-brother, his father's chosen heir and his first- 
born son by a high-ranking chiefess, but from court attendants. The 
father himself, once he is certain that the newcomer is his son, performs 
the role of the ideal father in all but one important respect. He does not 
choose him, instead of his half-brother, as his heir and successor. 

Unlike the Old World and Near Eastern myths, the Hawaiian narra- 
tives omit the warning against the child's birth, the child's abandon- 
ment, and its later revenge for the abandonment. In most Old World 
sagas, the hero is reared by neither of his natural parents; in the Hawai- 
ian, the child's natural mother always rears her child and lives with him 
among her kinfolk who assist in rearing him. 

Hawaiian narrators do not treat the natural father's absence as aban- 
donment since he left a name for the expected child and, in most cases, 
tokens of his rank to confirm the child's paternity. Narrators justify his 
absence as due to homesickness, the call of official duties, the need to 
prepare for the child's eventual arrival in his kingdom, and the search 
for a birth gift. Abandonment is thus glossed over. The child, on learn- 
ing who his natural father is, does not feel he was abandoned; instead, 
he is happy and eager to join him. 

Although the child questions who his natural father is, he never 
doubts that the woman rearing him is his real mother. The girl 
Laukiamanu, although seemingly hostile to her mother as well as to her 
mother's husband, takes for granted that she is living with her natural 
mother. The male father-seeker has a very protective mother who plays 
a large role in his life from the time of his birth until his departure. 
While he detaches himself from her by leaving, she may reappear in his 
life later. 'Umi's mother is a model. She rears him carefully, protests her 
husband's mistreatment of him, identifies his father, gives him the 
tokens, and tells him how to reach the court. She also gives him her 
younger brother, about 'Umi's age, as his squire. After King Liloa's 

The Foster Child in Hawaiian Myths and Customs 15 

death when 'Umi has to escape with his life from his hostile half- 
brother, now the king, he regresses and returns for shelter to her. How- 
ever, she warns him he is not safe with her but should seek safety in exile 
outside his half-brother's kingdom. After he becomes king, 'Umi's bond 
with his mother remains strong. He invites her to live at the court with 
her husband and her children born since his departure from home. 

A father-seeker's mother can also be extremely vindictive on his 
behalf. Nakif emaka and Nl'au each have a mother who enables them to 
journey to their father and who later takes revenge against the father 
for having initially rejected their son. Naku'emaka remains loyal to his 
father, who escapes death only through his assistance, but Nl'au's father 
is slain and transformed into a fish. 

Each mother at first evades telling her son of his father's identity, per- 
haps through reluctance to have him leave her, grow up, and face the 
dangers inherent in his ambition to claim his heritage. Each mother is 
determined — and is sometimes supported by her parents, grandparents, 
and ancestors — that his natural father shall acknowledge him and grant 
him the status they feel is due him. The son's journey is both a psycho- 
logical and a physical transition. It marks the death of his childhood 
and the start of his adult life. It removes him from his mother's side and 
her limited social environment to a larger arena and great danger. 

The case of the girl Laukiamanu presents a problem. Her mother 
appears hostile to her, but this may be due to the narrator forgetting to 
mention the father's tokens — both when he leaves and when the girl 
departs to find him. The mother never mentions the tokens and 
Laukiamanu leaves without them. The mother earlier seems reluctant 
to admit that the girl's father is the king that Bamboo told her about. 
She does tell the girl of her father's requirements as to the canoe and its 
gear that will identify her on her arrival. The headstrong and impatient 
girl will not listen, although the mother warns that she will suffer 
"untold agony." She then sends the girl to her own older relatives for 
transportation on a stretching bamboo to the king's cloudland. The 
tokens are mentioned for the first time when the grandmother flies in 
with them and saves the girl's life. At last she is accepted by her father. 
The apparent ambiguity of the mother may have resulted from jealousy 
of her daughter becoming the pampered chiefess in her former lover's 

The Hawaiian introduction of the conventional "Who is my father?" 
query into narratives about real kings like 'Umi and Kiha recalls Tho- 
mas Mann's statement that not only does the biographer seek to assimi- 
late his subject's life to a conventionalized form but that the subject may 

16 Pacific Studies, Vol. 10, No. 2— March 1987 

also see his life as the reanimation of myth, not, however, as "I am 

lik e " but as "I am — ." 18 The outline of these nine sagas illustrates the 

application of the pattern not only to semihistorical kings but to demi- 
gods and fictitious characters. Fantasy and reality have shaped the pat- 
tern. The many Hawaiian social customs relating to fosterage, adop- 
tion, parenthood, and rank have inspired children and even adults to 
believe that the fantasies in the sagas could become realities in their own 
life. These social customs are discussed next. 

A Basis for Fantasy in Social Reality 

Hawaiians surpassed other island peoples in the number of stories of 
mythic, semihistorical, and fictitious characters inquiring about their 
biological father's identity; they probably also surpassed them in elabo- 
rating a set of customs of adoption and fosterage not only of children 
but of adults. If, as Robert H. Lowie stated, "Oceania as a whole repre- 
sents a main center for adoption carried to unusual lengths," 19 then 
Hawaii may well be the heart of that center. 

Following is a summary of the pattern of adoption and fosterage of 
which there were (and still are) many variations and exceptions — 
enough to excite a young person's imagination to develop a latent fan- 
tasy of having very distinguished parents. In Hawaiian fantasies these 
ideal parents would be even more distinguished than the known natural 
or adoptive parents, for many, if not most, children, whether born of 
chiefs or commoners, were reared by other than their natural parents. 
Adoption followed class lines with commoners adopting children of 
commoners and chiefs those of chiefly rank. The natural parents either 
retained primary rights or surrendered them temporarily to surrogate 
parents. To prevent supernatural punishment with the child as the vic- 
tim, the natural parents had to fully approve the transfer, agree with 
the substitute parents as to whether the transfer was temporary or per- 
manent, not quarrel with them over the child, and if the transfer was 
supposed to be permanent, not take back the child unless the adoptive 
parents died. As the two sets of parents were usually kinfolk, members 
of the same extended family ('ohana) or close friends who usually lived 
near each other, the child was not cut off from its biological parents or 
extended family and was not kept ignorant of its origin. The adoption of 
the child was intended to reinforce existing alliances and create new 
ones beneficial to all concerned. 20 

The grandparents' claim to grandchildren took precedence over that 
of the natural parents, who had to get their consent to keep a child to 

The Foster Child in Hawaiian Myths and Customs 17 

rear for themselves. The firstborn, if a boy, customarily went to the 
paternal grandparents; a girl went to the maternal grandparents. 
According to Mary Kawena Pukui, "The whole feeling was that the first 
grandchild belonged [her italics] to the grandparents. The natural 
mother had the baby on a kind of 'loan' basis." Children born later 
might be adopted by the grandparents' siblings and lateral relatives and 
then by the parents' siblings and lateral relatives and friends. Adoptive 
parents reared and educated the child in a way appropriate to its birth 
order, social class, sex, and future career. An adopted child, particularly 
if firstborn, often became an indulged favorite but was responsible for 
becoming the family specialist in its traditional knowledge, supervising 
and caring for younger siblings, and eventually serving as family leader 
and counselor. 21 

English translators of Hawaiian texts often use the English terms 
"adoption" and "fosterage" interchangeably, and the line between them 
in Hawaiian custom sometimes appears to waver. But whether this is 
the result of different regional usage in the past or of modern Hawaiian 
interpretations is not clear. Two major Hawaiian terms relevant to the 
matter are ho'okama (to be discussed later) and hdnai. According to 
Pukui and Elbert, hdnai as a noun refers to an adopted or foster child. 22 
As an adjective it describes either the nurtured or the nurturing person. 
The child in question is keiki hdnai or he hdnai. As a verb hdnai means 
to feed, rear, nourish, and sustain. I find that the adoptive parent, if a 
commoner, is called makua hdnai, but if of chiefly rank, more properly, 
kahu hdnai. Pukui, emphasizing the permanency of the hdnai relation- 
ship, has stated that a child "is the hdnai of his permanent, adoptive 
parents" and the relationship is as permanent as that in modern legal 
adoption (ho'ohiki) . 23 The difference is that in modern legal adoption 
outside the family the child severs all ties with natural parents and other 
kin, whereas in hdnai he retains contact with his natural parents and 
'ohana. How generally Pukui's interpretation of hdnai as a permanent 
relationship would have been accepted in earlier times is uncertain. 
According to Pukui, a child cared for temporarily or part-time by foster 
parents was called luhi, and its natural parents had the right to reclaim 
it when they wished. Today some individuals reared for a long time by 
other than their natural parents either do not understand or remember 
whether they were luhi or hdnai of a relative, the foster child or the 
adopted child. 24 They also confuse the relationship with modern legal 
adoption and may seek legally to get a share of the adoptive parents' 

References to 'Umi's boyhood demonstrate the varied traditional uses 

18 Pacific Studies, Vol. 10, No. 2— March 1987 

of the term hanai. When 'Umi, who had been carefully brought up 
(hanai) by his natural mother, Akahiakuleana, asked her if the man 
who beat him was his only makua kane (literally, male parent) and if he 
had a different one, he either hoped or suspected that the man was not 
his luau'i makua kane ("true" or natural father) but either his makua 
kane hanai ("feeding" or adoptive father) or makua kane kolea (stepfa- 
ther) and that he was his keiki hanai ("feeding" or adopted son). 'Umi, 
listeners know, is not contemplating merely moving to a nearby rela- 
tive's or friend's house following the custom of commoners' children 
who, whether ill treated or not, roamed at will from one household to 
another. Because 'Umi already knew that he had another makua kane — 
his mother's younger brother 'Oma'okamau — it is clear that he is asking 
about a more mature makua kane. The term designates either the natu- 
ral father or the uncles and male cousins of both parents' generation. 

David Malo called 'Umi's stepfather makua kolea (literally, plover 
parent), a metaphor likening a stepfather to the migratory Pacific 
Golden Plover, which does not breed in Hawaii but winters there before 
flying back to the Arctic to nest and raise its young. Although N. B. 
Emerson commented that the metaphor is "used rightly without a 
laugh," one still hears it said jokingly. Emerson considered makua kolea 
"a very significant phrase" because "of the uncertainty of the parentage 
on the male side," as expressed in the proverb "One can be sure of the 
mother but not of the father." 25 

Proof of the identity of both parents was important to chiefs in tradi- 
tional culture, for chiefs constituted a caste-like class apart from com- 
moners. Long genealogies of both parents supported claims of a chief to 
a certain rank, within their own class. 26 If it was not definitely known 
which of two chiefs was a child's father, presumably either the one with 
the higher rank or the one more generally accepted was used in the for- 
mal genealogy. The highest ranking chiefs and chiefesses were regarded 
as the earthly representatives of their divine ancestors, and great care 
was taken that there be no doubt as to who the father of a firstborn 
child was. 

According to Abraham Fornander, a chief's rank, being determined 
by birth, did not decline if he lost possessions or influence, nor could it 
be raised by wealth and power. He might, however, through alliances 
by marriage and adoption "raise the rank of his children higher than his 
own, such as by marriage with a chiefess of higher rank than his own, 
marrying with a sister, or by their adoption into a family of higher rank 
than the father." 27 Ruling chiefs married several chiefesses for political 
reasons and to have high-ranking children. Kamehameha the Great is 

The Foster Child in Hawaiian Myths and Customs 19 

an example. Both the mother and the maternal aunt of his sacred wife 
Ke'opuolani, who bore Kamehameha II and III, considered Kameha- 
meha, like all chiefs from the island of Hawaii, of inferior rank; his mil- 
itary successes could not alter the fact in their minds. 

Paternity was often in doubt as there was much sexual freedom, even 
for sacred chiefs and chiefesses once they had produced an heir. Nine- 
teenth-century historians cite many examples of uncertain paternity 
among chiefs and chiefesses. The most famous example is 'Umi's descen- 
dant Kamehameha the Great, who had always thought that Chief 
Keoua of the island of Hawaii was his natural father until a few years 
after he conquered all the islands but Kauai. In S. M. Kamakau's 
account, Kamehameha, on hearing that Chief Kame'eiamoku was 
dying at Lahaina, Maui, went to him. Kame'eiamoku had attended 
Kamehameha since his boyhood and was one of his four great coun- 
selors who had engineered his rise to power and made him king. 
Kame'eiamoku kissed Kamehameha and said, "I have something to tell 
you: Ka-hekili was your father, you were not Keoua's son. Here are the 
tokens that you are the son of Ka-hekili." What the tokens were is not 
stated. Kamehameha commented that it was strange that his lifelong 
friend told him this now; had he told him earlier his "brothers" (Maui 
kinsmen) need not have died, for the rule could have been divided with 
them. His counselor explained that it was better to have conquered the 
islands because with one ruler over them there would be peace. 28 Kahe- 
kili, whom Fornander called "the reputed, if not the acknowledged 
father of Kamehameha I," was a very high chief who before his death in 
1794 ruled Maui, Molokai, Lanai, and Oahu. 29 Kamehameha had 
fought against him in invasions of Maui led by Keoua's half-brother 
Kalani'opu'u and later by himself. 

The old counselor's dying revelation sheds light on why Kahekili, on 
hearing of Kamehameha's birth on Hawaii, had sent his half-brothers, 
the twins Kame'eiamoku and Kamanawa, to serve the child as honored 
attendants (kahu), a role they faithfully played even against their rela- 
tives until their death. Kamakau commented that it was an ancient cus- 
tom "for the chief of one island to give a child to the chief of another 
island. This is the reason why Ka-hekili has often been called the father 
of Kamehameha, for chiefs of Hawaii and Maui were closely related, 
and this is why the twins Ka-me'e-ia-moku and Ka-manawa, who were 
the children of Ke-kau-like, ruling chief of Maui, were made tabu to 
live on Hawaii as associates for the child of Ka-hekili." 30 

The revelation, it is said, led Kamehameha to consider himself an alii 
po'olua (double-headed chief), but Kamakau's account does not support 

20 Pacific Studies, Vol. 10, No. 2— March 1987 

an implication by E. S. C. Handy and Mary K. Pukui that Kameha- 
meha and other Hawaiians thought that he had two biological sires, a 
phenomenon that they claimed was "believed to occur among a/ft." 31 
Among other examples of double-headed chiefs is one of Kamehameha's 
wives, Kaheiheimalie, who was said to be the daughter of either 
Ke'eaumoku (another of Kamehameha's great counselors) or perhaps 
Kanekoa. 32 Sometimes a chiefess was already pregnant by one husband 
when she became the wife of another in whose household her po'olua 
child was born. One such child was Ke'elikolani (Princess Ruth), Kame- 
hameha's great-granddaughter, born in 1826. Her mother, it seems, was 
pregnant by Chief Kahalai'a when she became the wife of Chief 
Kekuanao'a who is listed as the father of the child, said to be po'olua, 
"that is, a child of two fathers, which was considered a great honor by 
chiefs of that period." 33 Malo, explaining the term po'olua differently, 
stated that when a chiefess with a high chief as father but a lower- rank- 
ing chiefess as mother gave her child by her husband to another chief for 
adoption the child was called alii po'olua™ Sometimes a po'olua child 
resulted from deliberate planning, but whether only among commoners 
or also among chiefs is unclear. If a woman bore only stillbirths or no 
children her husband might approve of her having relations with 
another man because, it was believed, if she had a child by him then she 
and her husband would have children of their own after that. If a 
woman was uncertain which of two men had fathered her child both 
men might jointly accept it as theirs. The double-headed child then had 
more sets of relatives to care for it and the relatives had at the same time 
extended their alliances. 

According to Malo, "Women very often gave their children to men 
with whom they had illicit relations," and a chief who had children by 
secret amours might recognize some of the children, some not. One who 
was told of his chiefly ancestry, although the public might not always 
know of it, was called "a chief with an ancestry" because he knew and 
could prove his pedigree. A child who merely knew but could not prove 
his chiefly bond was a "clothes-rack chief" because he put on airs such 
as not permitting anyone to put their clothes on his shelf. 35 Malo did not 
state the rank, if any, of the women he referred to but presumably they 
were chiefesses but of lower rank than the chief. Usually if mates were 
of different rank their offspring became lesser chiefs in the ruler's 
court. 36 

Although the social classes were theoretically endogamous and caste- 
like, casual mating occurred across class lines, providing children of 
commoners with a social basis for nourishing the fantasy of being the 

The Foster Child in Hawaiian Myths and Customs 21 

offspring of a great chief. The mating might have been during a para- 
mount chiefs circuit of his domain with his retinue, or like 'Umi's royal 
father Liloa after his return from a journey to dedicate a heiau. Some- 
times a chief or a chiefess either at court or in the country took a fancy 
to an attractive commoner. Neither the commoner nor the commoner's 
spouse dared object although the ranking lover's spouse might. To pre- 
vent any offspring born from such liaisons claiming rank through a 
blood tie, chiefly families had them killed in infancy. It was particularly 
important that a king's children resulting from such liaisons be killed, 
either at the order of the king or his council of high chiefs, to forestall 
any claims to rank or, more importantly, rulership. 

King Liloa, it would seem, acted most irregularly in leaving tokens of 
his rank for a son by a woman of lowly birth, perhaps not even a mem- 
ber of the chiefly class. Customarily, the offspring of such a mating 
would have been sought out and killed to prevent what did happen — 
'Umi's arrival at court with the tokens, his acceptance by Liloa, and 
when Liloa died, his transformation into a dangerous competitor of his 
royal half-brother for power. 

If the liaison was unwittingly or deliberately with an attractive 
kauwa (member of a segregated, polluting caste), families made every 
effort to separate the couple, and if a child were born "it would be 
dashed to death against a rock," according to Malo. If a woman knew 
neither the name nor the lineage of her child's father, the child was 
called, according to Handy and Pukui, keiki a ka pueo (child of an owl), 
the term for one begotten by the roadside by an unidentified man. 37 

Kamakau has stated, "There is no country person who did not have a 
chiefly ancestor. The kauwa too had a few born of them [chiefs] who 
concealed their relationship on the side of the slave." 38 Sexual freedom 
was such that an ambitious boy living as a commoner had abundant 
cause to fantasize being related to a chief, and to dream of someday 
having his rank recognized and chiefly privileges accorded him. A chief 
from whom many claimed descent was Keawe, one of 'Umi's sons and 
successors, and, it will be recalled, the ruling chief that Paka'a served. 
Keawe, who had five (some say seven) aristocratic official wives, also 
had, according to Fornander, "numerous amours with women of low 
degree and with the daughters of the common people." As a result some 
genealogists "greatly blamed [him for] thereby impairing the purity of 
the aristocratic blood and giving rise to pretensions that in after ages 
. . . became difficult to disprove." Nonetheless, Fornander "found 
... no name or family claiming descent from him and setting up pre- 
tensions accordingly, unless they were actually and historically descend- 

22 Pacific Studies, Vol. 10, No. 2— March 1987 

ed from one of his five wives. . . ." 39 He was writing no doubt of mem- 
bers of the chiefly class, not of commoners who would talk among 
themselves of his amours. 

Nevertheless, legendary and historical accounts tell of chiefs engaging 
in battle for supremacy and fighting not only with weapons but with 
ritualized boasts of high birth and taunts about the other's low ancestry. 
After describing young Kamehameha's battles on Hawaii against a rival 
chief, Kamakau added: 

The strife between the chiefs took the form of denying each 
other's pure descent from a line of high chiefs. Each was well- 
versed in genealogical lines, oratory, and minute details in the 
histories of chiefs, their birthplaces, rules of government, and 
the signs and omens that revealed their rank as chiefs. Both 
sides also had composers of meles who chanted the names of 
ancestors, the high and godlike rank of their own chief, and the 
mean ancestry of the other. This form of controversy between 
two chiefs is well-known today and will be remembered for all 
time. 40 

A malicious person could start a rumor questioning an opponent's re- 
puted paternity, or a commoner might claim high birth through his 
father with the hope of rising in the world. Beck with wrote: 

But for genealogical purposes a wife's children were generally 
accepted as his own by the nominal husband unless the actual 
parent was in a position of advantage in rank and power which 
made him worth cultivating by an ambitious offspring. The 
journey of a first-born child of his mother to seek recognition of 
a highborn father in a distant land is hence a favorite theme of 
Hawaiian sagas and romances. 

The effect of such loose matrimonial relations in a land 
where inherited blood counted above all things in establishing 
the perquisites of rank is to be seen in the dual pattern of court 
genealogies, where an unbroken line of descent often depends 
upon the female when a male parent fails. The Keawe line 
from 'Umi is twice so preserved on the 'Ulu genealogy. Both 
genealogies for the Kalakaua family derive finally through the 
mother. 41 

Even if a youth proved that his natural father was an important chief 
or king and was accepted at court, he was handicapped by his mother's 

The Foster Child in Hawaiian Myths and Customs 23 

lowly birth. To belong to the highest echelon of chiefs and chiefesses, a 
person had to have not only a high-ranking father but a high-ranking 
mother. Traditionally a king chose his successor on the basis of his first- 
born son's mother's rank being superior to that of the king's other wives. 
However, he consulted other high chiefs to get their support for his deci- 
sion. If several chiefesses were of equally high rank, a king took all of 
them as wives to prevent their taking husbands whose children would 
compete on the basis of rank for power after his death. For the same 
reason his marriage to his sister or half-sister was desirable to insure the 
purity of the line and produce a child higher in rank than both parents. 
If the most eligible son was considered incapable by the ruler and his 
council he was passed over for a better qualified younger son or some- 
one outside the immediate family. If the sons were all by women of infe- 
rior chiefly rank, the daughter of the highest ranking wife would be the 
king's successor, but only two such cases are known. 42 

When 'Umi's foster-child fantasy became reality and his royal father 
acknowledged him, his mother's humble status, whether that of com- 
moner or very low-ranking chiefess, was debated by friends and ene- 
mies. To jealous Hakau, 'Umi's royal half-brother and their father's 
heir, 'Umi's mother was even lower than a commoner. After calling her 
a wahine kauwa ("slave" or outcast woman), he told Liloa that he had a 
kauwa for a son, a shocking insult to Liloa because ordinarily neither 
chief nor commoner would sleep with an untouchable. 'Umi's defenders 
claimed that his mother and Liloa were cousins through a common 
highborn ancestor, a relationship that Liloa himself had ascertained at 
their romantic meeting by asking "Who is your father?" Also, when he 
met 'Umi he asked kindly after his mother and sent presents to her and 
her husband. According to Kamakau, 'Umi's mother's genealogy 
"shows that they [her kin] had fallen very low." 43 

'Umi's enemies called him noanoa (commoner), keiki kapa alii (part 
chief), as well as kukae popolo, keiki lepo popolo, and lepolepo, which 
metaphorically refer to a chief whose mother was said to be a com- 
moner but literally refers to excrement from eating popolo greens. 44 
These terms, said Kamakau, "might have been right or perhaps they 
were not, but he was victorious and ruled the kingdom." 45 Nonetheless, 
whatever power and glory a chief achieved through courage and ambi- 
tion could not, as Fornander pointed out, raise his rank, although he 
could raise that of his children. 

When 'Umi became king, a priest named Kaleioku (or Ka'oleioku), 
who had adopted him when he was in exile and was responsible for his 
deposing Hakau, advised him how to rule and whom to marry to have 
high-ranking children in order to dampen the criticism of the high 

24 Pacific Studies, Vol. 10, No. 2— March 1987 

chiefs that his mother's lowly birth did not warrant his becoming king. 
Thereupon, 'Umi took several blue-blooded chiefesses as his official 
wives, most importantly his half-sister Kapukini-a-Liloa, who bore two 
of his successors (although she thought 'Umi's rank slight). One of the 
sons was Keawe the Great. 'Umi may also have married Pinea, Hakau's 
daughter named for her mother. More is known about his wife Pi'ikea, 
daughter of King Pi'ilani and Queen La'ielohelohe of Maui and sister of 
Pi'ilani's chosen successor and firstborn named Lono-a-Pi'i and of Kiha- 
a-Pi'ilani. The latter, with 'Umi's military support, was to depose and 
kill Lono to become king. 'Umi also mated with country women by 
whom he had numerous children; it is said that if any commoner on the 
island of Hawaii declared that 'Umi-a-Liloa was not his ancestor it was 
through ignorance of his ancestry. 46 This recalls Fornander's remark 
that anyone who did not claim descent from Keawe did not want to 
do so. 

King Pi'ilani, it should be added, had been called kukae popolo and 
kukae paoa (stinking excrement) for the same reason as 'Umi. His 
mother's pedigree was not even remembered or perhaps conveniently 
forgotten, but Fornander charitably thought that she "was probably 
some Maui chiefess." 47 To have children of higher rank than himself, 
Pi'ilani married his paternal cousin, highborn La'ielohelohe of the 
important Kalona family on Oahu where she preferred to live and 
where she reared Kiha. 

Events relating to 'Umi's and Kiha's overcoming the handicap of 
being a king's younger son illustrate another type of adoption besides 
that of children. A young adult might be adopted as a protege by a per- 
son older than himself, who would take him into his household to feed, 
protect, train, and counsel so that the careers of both the adopted ward 
and the adoptive parent would be advanced. 'Umi and Kiha each 
became a protege of an older man, when, after their royal father's 
death, they had to flee and live incognito in exile as commoners to 
escape death at the hands of the new king, their abusive and insulting 
older brother, who had always resented having his prerogatives threat- 
ened or actually usurped. 

Chroniclers differ about exactly how 'Umi and Kiha eventually 
became king, but what follows provides a notion of the events and their 
adoptive father's role in them. 

Each chief had the good fortune to be recognized by one or more wily 
and ambitious priests who, hoping later to be rewarded, adopted the 
exiled youth in order to depose the hated new king and replace him with 
his junior. Kaleioku, a priest and lower in rank than 'Umi, made 'Umi 

The Foster Child in Hawaiian Myths and Customs 25 

his haku (lord, master) and alii (chief) and made himself 'Umi's kahu 
hanai (adoptive father). He took 'Umi, his commoner wife, and their 
children into his household to care (hanai) for them. He also began to 
hanai (feed, support) 'Umi's loyal companions and others to build an 
army against Hakau. The proverb Hanai kanaka, hiki ke ho'ounauna, 
meaning "Feed human beings, for they can be sent on errands," refers to 
the benefits of treating an adopted child well. 48 Perhaps Kaleioku had 
the proverb in mind. Reciprocally, the army in serving Kaleioku and 
'Umi could be said to hanai them. When 'Umi became king he 
rewarded Kaleioku with lands and the position of high priest and chief 
counselor, and brought his mother and her younger children to live at 
court. Nothing is said of his two children by his commoner wife or of 
their ever trying to claim 'Umi as father and benefit from the relation- 

In exile on Maui with his highborn wife, Kumaka, Kiha was adopted 
by Chief Kahuakole, who advised him to put aside Kumaka and marry 
the Hana chiefs favorite daughter Koleamoku in order to get her father, 
Chief Ho'olae-makua, to support him against Lono-a-Pi'i. 49 Ho'olae dis- 
owned Kolea, saying that she should have first married Lono to whom 
she was tabooed and that after that she could have taken another hus- 
band. When the couple's son Kauhiokalani was born, Kiha sent Kolea 
with the child to ask Ho'olae for certain lands. This request for strategic 
lands made Ho'olae realize that his hated son-in-law was not a com- 
moner but Chief Kiha who was planning to rebel against King Lono-a- 
Pi'i, to whom he was loyal. He refused the request but kept the firstborn 
son of Kolea and Kiha to hanai. Kahu'akole then sent Kiha to his 
brother-in-law 'Umi for help. Urged by Pi'ikea and approved by 
Kaleioku, 'Umi invaded and conquered Maui. Among the dead were 
Lono-a-Pi'i and Ho'olae, but Kiha ordered that Kolea and Kauhioka- 
lani be spared. One assumes that Kiha then took his son to rear, but he is 
not mentioned again until after Kiha's death when he reappears serving 
his half-brother King Kamalalawalu, son of Kiha and Kumaka (whom 
Kiha had taken back). 50 After 'Umi and Kiha had apportioned the con- 
quered lands, 'Umi returned to Hawaii, but sent Kiha one of his and 
Pi'ikea's two sons. Kiha, however, despised his sister's children because 
of their father 'Umi's humble origin on his mother's side. (Yet 'Umi had 
just made him king!) Kiha treated the hanai with contempt, and when 
he killed the youth's favorite attendant (kahu), the youth died of grief. 51 

'Umi had other problems arising from his wife Pi'ikea's family. Her 
mother, La'ielohelohe, apparently did not scorn Pi'ikea's and 'Umi's 
children, as it is said that she asked for the firstborn and 'Umi promised 

26 Pacific Studies, Vol. 10, No. 2— March 1987 

it. He broke that promise, but promised the next, and so on without 
ever intending to honor his promise. He made these false promises, 
according to one chronicler, because he had once vowed that, contrary 
to custom, he would not permit someone else to rear any child of his. 
Finally, when Pi'ikea was pregnant again, La'ie sent her supernatural 
ancestresses, "grandmothers," to Hawaii once more. This time they had 
the gods strike 'Umi's people dead at night. When he heard of the 
deaths, he foolishly went out to fight the gods. Meanwhile the grand- 
mothers, acting as midwives, obtained the newborn and carried it to 
Oahu. Thus, says a chronicler, it came into La'ie's possession, that is, 
was adopted by her. 52 

According to Kamakau, "It was regarded as a great honor for a chief 
to be reared by his grandparents, and for the chiefs to rear their chil- 
dren's children. This made the chiefs beloved." 53 A grandchild, particu- 
larly if it was the mother's firstborn, was much indulged. Kekauluohi, 
daughter and firstborn of Kaheiheimalie and Kala'i-mamahu (Kameha- 
meha's younger half-brother), was reared by her maternal grandpar- 
ents, Namahana and Ke'eaumoku, who "fondled her as if she were a 
feather lei from the precious mamo bird." She was "a favorite above all 
the other grandchildren," and was also the favorite of the uncles and 
cousins of her aunt Ka'ahumanu, her mother's older sister and one of 
Kamehameha's wives. Kekauluohi was looked on as the family head, 
and her father's own trusted kahu and the latter's kin were her care- 

When Kamehameha took Kaheiheimalie from his half-brother to be 
one of his wives, the child probably remained with her maternal grand- 
parents because Kaheiheimalie had refused to let Kamehameha adopt 
her at her birth because she loved Kala'i-mamahu and wanted to rear 
their child. Kekauluohi later became Kamehameha's youngest wife, co- 
wife (punalua) with her mother, her mother's sister, and other high- 
ranking chiefesses. After Kamehameha's death his son Liholiho (Kame- 
hameha II) took her as one of his wives but later gave her to his friend 
Kana'ina. By Kana'ina she had a son Lunalilo who succeeded Kameha- 
meha V as king. She and Kana'ina were the adoptive parents (kahu 
hanai) not only of Kalama, who became the wife of Kauikeaouli 
(Kamehameha III), but of the royal couple's second son. 54 

At the time that her aunt Ka'ahumanu was caring for Princess Ruth, 
Kekauluohi helped her rear Ka'ahumanu's grandnephew and her own 
nephew, David Kamehameha, firstborn of Klna u and Kekuanaoa. 
Klna'u was Kekauluohi's half-sister and like her had been one of Liholi- 
ho's wives. David's birth had helped reconcile Ka'ahumanu to Klna'u's 

The Foster Child in Hawaiian Myths and Customs 27 

refusal to marry Kauikeaouli (Kamehameha III) in accord with the 
wish of Kamehameha the Great that Klna'u and Kamamalu, his daugh- 
ters by Kaheiheimalie, marry his sons by Ke'opuolani, his sacred wife, 
to continue his line. 55 Kamamalu, also one of the wives of Liholiho 
(Kamehameha II), had died with him in London. 

Kekauluohi's adoption by her grandparents and her subsequent 
adoption of others is only one of many examples concerning the highest 
nobility and royalty. With rare exceptions, a royal or other highborn 
child was given by its parents at birth or soon after to another high chief 
or chiefess, usually related in some way, who became its kahu hanai. 
The term kahu, with or without a descriptive adjective, refers only to 
the guardian or attendant of a child (or adult) of status, not to that of a 
commoner. A ruling chief with no children of his own might have his 
hanai child succeed him if the child was old enough. People praised a 
wealthy chief or chiefess who reared a poor chief's child. More than one 
wealthy kahu hanai made his adopted child his heir. 56 A minor chief 
who reared a high-ranking chief's child benefitted socially and politi- 
cally, as did his children. For the same reason a high chief might rear a 
child of a chief only a degree higher than himself. Royalty also found it 
politically advantageous to rear their relatives' children and grand- 
children and to have relatives rear theirs. 

Each of Kamehameha's two sons by his sacred wife Ke'opuolani (his 
niece who outranked him) was given to a kahu hanai. However, in 1797 
the firstborn, Liholiho, was taken from his original caretaker after his 
maternal grandmother reported that the infant (then about six months 
old) was not getting the right diet and the wet nurses were careless. 57 A 
chiefly child had several chiefesses as wet nurses because of numerous 
taboos relating to their duties that if violated would supernaturally 
harm and pollute the sacred child. 58 Kamehameha then gave Liholiho 
to Ka'ahumanu, his favorite wife and cousin, to rear, and thereby had 
this son with him much more than was customary. 

Chiefs jockeyed for the honor of being a highborn child's kahu hanai, 
and the chosen chief was present at the birth, ready with wet nurses 
who were chiefesses in his family. After several chiefs had begged 
Ke'opuolani for her next child, she promised it to her kahu, Kuakini, 
brother of Ka'ahumanu and Kaheiheimalie. When the child was born 
in 1814 (the date is uncertain), Kuakini, thinking it stillborn, rejected 
it. Kaikio'ewa, one of the king's cousins and brothers-in-law, arrived in 
time for his accompanying high priest Kapihe to get the newborn to 
breathe and was made Kauikeaouli's kahu hanai. When Kaikio'ewa 
was appointed governor of Kauai to replace Kahalai'a, the latter was 

28 Pacific Studies, Vol. 10, No. 2— March 1987 

appointed the boy's kahu hanai to console him for being removed as 
governor. When Kauikeaouli became King Kamehameha III in 1825, 
succeeding Liholiho, he was only ten or eleven years old. He had a suc- 
cession of guardians with various responsibilities, but Kaikio'ewa con- 
tinued as a major kahu until his death in 1839, and would stop at noth- 
ing, even at the risk of his life, to further the well-being of the king and 
the kingdom. 59 Kaikio'ewa was also the adoptive father of Moses 
Kekuaiwa Kamehameha, second son of Klna'u and Kekuanao'a, and 
made him his heir. 

A kahu hanai who reared his hanai child from infancy was responsi- 
ble for its personal care and upbringing and supervised other kahu, his 
relatives, who assisted in caring for it. As usual the natural parents 
maintained close contact with the adoptive parents and the child, but 
where royal children were concerned the situation of a kahu hanai dif- 
fered. He obviously did not have primary rights to a child who was in 
the line of succession, and he could be relieved of his position or have it 
modified as the natural parents and other advisors of the king and 
queen decided. Kamehameha the Great, before his death in 1819, 
appointed a kahu-nui (or kahu alii), a supreme kahu, who was over all 
kahu hanai and other kahu associated with the royal family. Kameha- 
meha appointed the chief he most trusted, Ulumaheihei Hoapili, 
Kame'eiamoku's son, as the sacred family's chief guardian. He also 
entrusted to him the hiding of his bones after death. Further, he gave 
him Ke'opuolani as his wife (after her death in 1823, Hoapili married 
Kaheiheimalie). Until his death in 1840, Hoapili, who had great 
authority, was constantly consulted by members of the royal family and 
by other chiefs and chief esses in matters relating to the royal family, 
upon whom the welfare of the kingdom depended. It was to him, for 
instance, that Klna'u and Kekauluohi, the young king's "foster moth- 
ers," went with others to discuss the errant behavior of Kauikeaouli 
after Ka'ahumanu's death. 60 

'Umi's refusal to give his mother-in-law his children to rear shows 
that Hawaiian parents did not always give up their children readily to 
others to rear. The experiences of Ke'opuolani and Klna'u further illus- 
trate the point. Ke'opuolani "wept when they [her two sons] were taken 
from her to be brought up by other chiefs and chiefesses. The mother 
yielded until it came to the last child and this one's rearing she would 
not give to another." The child was Princess Nahi'ena'ena, born in 
1815. 61 

Each of Klna'u's and Kekuanao'a's four sons had been adopted- 
David by Ka'ahumanu, Moses by Kaikio'ewa, Lot (later Kamehameha 

The Foster Child in Hawaiian Myths and Customs 29 

V) by Nahi'ena'ena, and Alexander (later Kamehameha IV) by Kauike- 
aouli, who made him his heir. When Klna'u's last child, Victoria Kama- 
malu, was born in 1838, she refused her maternal uncle Kuakini's 
request to take the child to the island of Hawaii to rear. Defying custom, 
she herself nursed it and her adopted daughter Pauahi, but made John 
Papa I'i and his wife Sarai her child's kahu. Their covenant, as Pi put it, 
was that he and Sarai would be the ones to carry the child, soothe it, 
and hold it on their laps. 62 Klna'u at the time was Premier of the king- 
dom, and Pi, a chief with important relatives and associated with the 
royal court since boyhood, had served both Liholiho and Kauikeaouli 
and was now Klna'u's secretary and adviser. When Klna'u died the fol- 
lowing year, Pi and Sarai became Kamamalu's foster parents under 
Kekuanao'a's supervision. 

A kahu hdnai was sometimes called an 'uhd (lap), which expresses the 
intimacy of guardian and ward. The term 'uhd was also used in other 
contexts. Some chiefesses were said to have tabooed laps. They were 
unable to rear either their own or adopted children because their auma- 
kua, or personal god, would make the children waste away or become 
crippled. The children thrived if adopted by others. The devoted auma- 
kua, people said, did not want the chiefess soiled by a baby. If she could 
rear daughters but not sons it was because her aumakua was male and, 
like a husband, jealous of sons. If a child accidentally wet a sacred 
chief's lap, he might have it killed or take it as a foster child. The par- 
ents, who would be the chief's personal attendants, would not be 
responsible if the chief himself picked up the baby and got wet; how- 
ever, he could still keep the child. Usually only a royal child had the 
right to sit on its royal parent's lap. 'Umi, in asserting his claim to King 
Liloa as his father, escaped the king's guards to sit on his lap. 63 

Despite its many guardians, a chiefly child was subject to hazards, 
which could inspire a lowborn child with fantasies of exalted lineage. 
Chiefs sometimes hid in the country with their children and lived incog- 
nito as commoners because of the abundance of food there, a desire to 
obscure the family's history as captives, weariness of observing their 
own taboos, and fear of the ruling chief who was perhaps a usurper. 64 
For example, after his power as a ruling chief on Hawaii was seized by 
Kamai'ole, Kanipahu hid his two sons (half-brothers) Kalapana and 
Kalahuimoku with trusted friends in secluded Waimanu Valley, Hama- 
kua district, where Kamai'ole would not find and kill them. Kalapana's 
mother and perhaps Kalahuimoku's remained with the boys. Years later 
when Kanipahu, who had been living as a commoner on Molokai, was 
asked to overthrow Kamai'ole he refused because he was ashamed of the 

30 Pacific Studies, Vol. 10, No. 2— March 1987 

calloused humps on his neck from carrying heavy burdens. He sent the 
messenger on to Kalapana, his chosen successor. Kalapana fought and 
killed Kamai'ole to become the ruling chief. His descendants included 
Liloa, 'Umi, and Kamehameha; 'Umi's mother was one of Kalapana's 
half-brother's descendants. 65 

Other hazards to children of rank came not only from shifting politi- 
cal conditions but from kidnapping, abduction, murder, abandonment, 
substitution, and mysterious disappearance. Kamehameha, a valuable 
political pawn, was kidnapped at birth and an attempt was made to 
abduct him when he was probably in his teens. Pregnant Keku'iapoiwa, 
a high-ranking chiefess and wife of Keoua, one of the war chiefs of Ala- 
pa'i-nui, a ruling chief on Hawaii, was in an expedition led by Alapa'i 
(Keoua's adoptive father) against the ruler of Maui. One cold and 
stormy night while the fleet was still at harbor before leaving for Maui, 
Keku'iapoiwa gave birth to Kamehameha unattended while her guards 
and the chiefs slept. To be alone was most unusual because in any crisis 
of illness, death, or birth, kinfolk and others gathered to give psycholog- 
ical and practical assistance. In the case of a highborn chiefess protec- 
tion was necessary to prevent a substitution or kidnapping. Nae'ole, a 
Kohala chief who wanted to be the kahu of Keku'iapoiwa's child and 
was waiting for the birth, made a hole in the thatch, snatched the new- 
born, and escaped with it. After its whereabouts were discovered some 
time later, the kidnapper and his younger sister were permitted, strange 
to say, to keep the child for about five years. 66 Then Alapa'i-nui took 
him for his favorite wife, Keaka, to rear. Kamehameha became very 
fond of his foster mother and her male cousin (an older relative of John 
Papa Fi) who helped to rear him. 

Earlier, Alapa'i had adopted Kamehameha's father Keoua and his 
half-brother Kalani'opu'u, and had taken their very high-ranking 
mother, his cousin, as a wife. Both natural fathers were dead; Kala- 
ni'opu'u's father had been slain by his half-brother, Keoua's father, and 
the latter had died in battle against Alapa'i. Historians ask, Did Alapa'i 
adopt the fatherless youths through kindness, or "to prevent them from 
hatching treason and revolt in the provinces," or both? The half- 
brothers, after becoming Alapa'i's war chiefs, distrusted him, feeling 
that he had no real regard for them. When Keoua died after a lingering 
illness, it was rumored that Alapa'i's sorcery had killed him. Fornander 
has rejected this black view of Alapa'i and considered Kalaniopu'u's 
next action ill-advised. 67 

According to Pi, Kalani'opu'u was mindful of Keoua's request that he 
look after Kamehameha, "who would have no other father to care for 

The Foster Child in Hawaiian Myths and Customs 31 

him." Fornander and Kamakau describe how he later risked his life 
attempting to abduct Kamehameha. He failed, and indecisive warfare 
against Alapa'i erupted. 68 Anticipating another attempt at abduction 
and control of Kamehameha, Alapa'i before his death warned Keawe- 
'opala, his son by Keaka, not to let Kalani'opu'u take Kamehameha 
away. When Kamehameha was about twenty and both Alapa'i and 
Kamehameha's natural mother were dead, Kalani'opu'u became his 
guardian and killed Keawe'opala in battle to become the ruling chief 
with Kamehameha at his side. 

There is a famous example of an adoptive father's treachery against 
his adopted son. Kahahana, an Oahu taboo chief, was adopted by King 
Kahekili of Maui as an infant. In 1773 when a council of Oahu chiefs 
selected Kahahana as their new king, crafty Kahekili, eager to capture 
Oahu for himself, used various strategems. The first included asking 
Kahahana to give him certain strategic Oahu lands as a reward for 
bringing him up. The Oahu council of chiefs refused the request. 
Although the faithful but gullible Kahahana fought with Kahekili 
against Kalani'opu'u, the Maui king continued to plot and finally 
created a situation that in 1783 enabled him to invade and conquer 
Oahu. Kahahana, who was captured after about two years in hiding, 
was murdered and sacrificed by his adoptive father at a heiau in 
Waikiki. 69 

Perhaps Kamehameha's mother's experience in having her newborn 
kidnapped led Ke'opuolani to warn Kuakini, selected as her next child's 
kahu hanai, to be present at the birth to prevent someone else gaining 
possession of it. Substitution of one newborn for another was also a pos- 
sibility. Ke'opuolani herself was almost replaced by another, although 
not at birth. She was so sacred that when she was with her wet nurse 
(kahu hanai waiu), anyone who dared approach or touch her would be 
burned to death. Yet her guardians, thinking her homely and puny, 
decided to substitute their own healthy daughter for her. They were 
thwarted, it is said, by a dog which entered where the daughter was 
sleeping and bit off the fingers of one hand. Kamakau commented, 
"The servant might have been the chiefess had not God willed it other- 
wise." 70 

Another instance of the unreliability of some caretakers concerns the 
taboo chiefess Kapi'olani, Kame'eiamoku's granddaughter and Uluma- 
heihei Hoapili's niece. When she was probably about five years old she 
was abandoned and almost died. Because Kamehameha was fighting 
the chiefs of Hilo where Kapi'olani lived, her caretakers (kahu malama) 
fled with her into the forest, but "tossed her into a clump of ferns . . . 

32 Pacific Studies, Vol. 10, No. 2— March 1987 

because her weight retarded them when danger was near." A man 
named Ho'omi'i, hearing the child's wailing, investigated, recognized 
Kapi'olani, and "picked her up and ran with sorrow for her in his heart 
. she might have died. If an enemy had found her, then she would 
have been killed." 71 Ironically, among the meanings of malama are 
fidelity and loyalty. 

According to Kamakau, "Some chiefs hid their children in the back- 
woods, and brought them up as commoners, and some children ran 
away into the back country and became countrymen." 72 How, why, and 
when Kai-ehu, Kame'eiarnoku's son by an unidentified mother, disap- 
peared has not been recorded. He was "born at the time of the battle of 
lao Valley" (1790) and vanished. Kame'eiamoku had his son Ulumahei- 
hei Hoapili search for his younger brother, "but no one knew where he 
was staying." Hoapili did not locate him until 1834 (if Kamakau's date is 
correct), and by that time Kame'eiamoku had been dead more than 
thirty years, Hoapili was nearly sixty, and the missing son was forty- 
four. Kaiehu "had been brought up in the country under the name of 
Ka-puni-'ai," and when he was found he was "peddling fish at Ka'ana- 
pali for some back countrymen who had no idea whose child he was." 
Unfortunately, nothing more is said about Kaiehu. Of the child's disap- 
pearance, Kamakau laconically remarked, "This is why the chiefs 
appointed a number of kahu to watch over a chief's child." One won- 
ders if Kaiehu ever wondered if he were a chief's child and if his case led 
some Maui commoners to imagine that they too might be the long-lost 
sons of high chiefs. 

Kaiehu presumably had become a waif and like some waifs who sur- 
vived had perhaps been adopted and reared as a servant to the adoptive 
family. A waif's treatment differed from that of a child adopted from a 
relative or friend, who might become such a pet that food would be 
dribbled into its mouth to prevent its choking on lumps, and who might 
be beautified to be exhibited in a rural beauty contest. 73 

A Ka'u family's ancestress is said to have been adopted as a child by 
her older maternal aunt who so neglected her that she ran away. She 
was found by an elderly couple who lovingly cared for her. Her natural 
parents knew nothing of her whereabouts until a prophetic dream led to 
a reunion just as the girl was about to be married. The bride at this time 
decreed that hereafter in her family only a younger sibling, not an older 
one, would be permitted to adopt a niece or nephew. 74 

Besides biological relationships, Hawaiians, as Handy and Pukui 
point out, have "three secondary categories of relationship whose basis 
is social rather than biological." They are hanai (the "feeding" relation- 

The Foster Child in Hawaiian Myths and Customs 33 

ship), marriage, and ho'okama. The latter, however, is only one of sev- 
eral structured and recognized types of friendship that constitute infor- 
mal adoption. Ho'okama means literally "to 'make' a child," and 
figuratively, "to adopt a child in friendship." The adopted and adopter 
are not biologically related or, at least, not close enough to be a consid- 
eration. The ho'okama relationship, which is "quite different from the 
'feeding' relationship," may be between persons of the same or opposite 
sex and of different ages. An adult may adopt a child, an elderly adult 
may adopt a younger adult. The relationship is based on "mutual affec- 
tion and agreement, at first tacit, then unobtrusively discussed, be- 
tween the child and the older person," and involves "love, respect, and 
courtesy, but not necessarily responsibility of any sort, and rarely a 
change of residence." The adopted individual is regarded as a member 
of the adoptive parent's family. If one party is indifferent to a tentative 
ho'okama or the parents strongly object, the matter is dropped. The 
adoptive parent, the one who initiated the relationship, is called makua 
ho'okama (parent making-child); an adopted son is keiki ho'okama; an 
adopted daughter is kaikamahine ho'okama. The relationship is some- 
what comparable to that between a godparent and a godchild. 75 In its 
broadest traditional sense, keiki ho'okama is how a chief in his role of 
makua ho'okama regards all those who are dependent on him. 

Variations of ho'okama are evident in an example from the 'Umi saga 
with regard to the rapidity of the development of the relationship, the 
change of residence, the deep responsibility between the adopter and 
the adoptee for each other, and the junior status of the adoptee in rela- 
tion to the adopter. 

Ten-year-old 'Umi, en route to his royal father, adopted three boys as 
his keiki ho'okama. For two of them, Pi'imaiwa'a and Ko'i, he first 
asked their own and their parents' permission, and they also asked their 
parents' permission. According to Kamakau, Pi'imaiwa'a's parents, rec- 
ognizing 'Umi as a chief because of the royal tokens, gave him their son 
"to live or to die in his service." Pi'imaiwa'a joked that 'Umi had a son 
now who had grown up for him in one day. Versions differ as to the 
third boy, 'Oma'okamau. According to Fornander, the boy was 'Umi's 
maternal uncle (makua kane), but the narrator later refers to him as a 
keiki ho'okama. In Kamakau's version, 'Umi and the boy were not 
biologically related and 'Umi had made him his keiki ho'okama long 
before leaving home with him; the adoption had been one more cause of 
'Umi's stepfather's anger toward him. 76 

Thomas Thrum states, "Keiki hookama, lit. adopted child, in this 
case is more that of a sworn boon companion, as they were lads together 

34 Pacific Studies, Vol. 10, No. 2— March 1987 

and in no sense as father and son. It illustrates a custom of companion- 
ship in expectation of sharing in the honors and good things of life. A 
close attendant, not a menial servant." 77 When 'Umi became king he did 
reward his loyal companions and he chose Ko'i to hide his bones when 
he died. Although the boys were contemporaries and boon companions, 
'Umi, who outranked them, was their lord (haku) as well as makua 

Only 'Umi among the father-seekers of the sagas had keiki ho'okama, 
but Nl'au and Laukiamanu had each acquired an intimate friend 
(aikane) of the same sex in the royal father's cloudland before the father 
recognized his offspring. In each case the father-seeker outranked the 
aikane. In Nl'au's case, a boy named Uhu'ula (Red parrot fish) who 
admired Nl'au's skill in games asked to become his aikane, to which he 
agreed. Whether it was Laukiamanu or her unnamed friend who first 
suggested that they become aikane is not stated. When the king's guards 
burned both Nl'au and Uhu'ula to death, Nl'au's ancestors restored him 
to life but transformed his friend into a red parrot fish. When Laukia- 
manu was thrown into a pigpen, she refused her friend's request to join 
her because she needed her to bring food. Later, when Laukiamanu, 
recognized by her father, was set free, she made her aikane a high 
chiefess and had her live with her. In each case a strong bond of affec- 
tion and loyalty existed between the two aikane, with the junior friend 
dying or willing to die if necessary. Only Laukiamanu's friend survived 
to be rewarded. 

The aikane relationship as described here is that of an intimate (but 
not homosexual) friendship between two members of the same sex, 
never between members of the opposite sex. The relationship is also 
called hale aikane (house friend) because of the mutual hospitality of 
the friends and their families. To each friend the other's relatives of the 
parental generation are inoa makua (parents-in-name), who care for 
their offspring's friend as they would blood kin. Descendants of two 
aikane sometimes continue to feel the link. 78 The relationship of the two 
friends is comparable to Old World blood brotherhood (and sisterhood), 
but no accompanying ritual has been reported for the Hawaiian rela- 

That the aikane relationship is a kind of informal adoption is expli- 
citly stated by Fornander: 

Among the members of the Aha-Alii [Congregation of Chiefs, 
descended on either the Ulu or Nanaulu line] it was not unusual 
that two young men adopted each other in weal or woe at all 

The Foster Child in Hawaiian Myths and Customs 35 

hazards, even that of life itself; and if in after life these two 
found themselves, in war time, in opposing ranks, and one was 
taken prisoner, his life was invariably spared if he could find 
means to make himself known to his foster-brother on the oppo- 
site side, who was bound to obtain it from the captor or the 
commanding chief. And there is no instance on record in all the 
legends and traditions that this singular friendship ever made 
default. 79 

Another custom is that described by Handy and Pukui of "making" a 
spouse. This is an adoptive platonic marital relationship, a kind of hon- 
orary marriage, and, like ho'okama and aikdne, it is a structured and 
recognized friendship. According to Handy and Pukui, 

A boy or man may take a great fancy to a girl or woman, mar- 
ried or unmarried. He tells her or his parents that he wants her 
as his "adoptive wife" or wahine ho'owahine [woman made- 
wife]. This does not imply having the sexual husband- wife rela- 
tionship, but a sort of brother-sister relationship. . . . Some- 
times a girl suggests that a certain man or boy become her kdne 
ho'okdne [man made-husband]. 80 

The girl or boy chosen may be a mere child while the proposed partner 
is a mature adult. Once the agreement has been made, the family and 
relatives of the individual who first suggested the relationship may 
cement it with a feast, for which a small pig is roasted. That any such 
feast formally concludes a ho'okama or aikdne agreement has not been 

Each partner has, in effect, a life-long friend of the opposite sex, an 
honorary spouse who is recognized as such by their families and friends. 
They are loyal, devoted, affectionate, and companionable without any 
sexual or economic involvement, but they exchange gifts. A poor but 
industrious man, unable to support a wife because of family obliga- 
tions, may become a girl's ho'okdne. From then on he brings whatever 
presents he can to her and if she marries to her husband as well. He and 
her husband treat each other as brothers and her husband regards him 
as a punalua, here used to refer to a sister's husband. The punalua will 
be as fond of his honorary spouse's children as if they were his own. 81 

Another form of platonic relationship is the reverse of that just 
described in that established sexual partners temporarily or per- 
manently adopt each other as brother and sister and use appropriate 

36 Pacific Studies, Vol. 10, No. 2— March 1987 

kinship terms to each other. Ho'okaikuahine refers to a man "making" a 
sister of his wife, who then calls him kaikundne (brother) and he calls 
her kaikuahine (sister). Kiha and Chiefess Kumaka, who had fled into 
exile together to escape from King Lono-a-Pi'i, had this relationship 
temporarily. Kumaka, according to Thrum, had been Kiha's "compan- 
ion in his trials and tribulations, even in those that might mean death. 
He made a sister of his wife." And called her sister. This was because his 
adoptive father, Kahu'akole, believed that if Kiha put aside Kumaka 
and married Koleamoku, the latter's powerful father would side with 
Kiha in his plot to depose Lono-a-Pi'i. Apparently Kahu'akole thought 
it more effective politically to keep Kiha's identity as a chief secret and 
not have Koleamoku and Kumaka live together as punalua, here mean- 
ing two wives sharing a husband. Because the plan failed, Kahu'akole 
told Kiha to leave Koleamoku and go back to Kumaka. When Kiha 
became king, Kumaka bore Kamalalawalu, who, when Kiha died, 
became king. 82 

The general features of the Hawaiian foster-child fantasy, as ex- 
pressed in the hero sagas, resemble those elements of Near Eastern and 
European sagas that psychoanalysts have described as part of a male or 
female child's normal psychological development toward maturity. 
(Such sagas, it should be noted, occur in other parts of the world besides 
those noted here.) Hawaiian social conditions under which a latent fos- 
ter-child fantasy could emerge and flourish differed, however, from 
those in the Near East and Euramerica. As this survey has shown, the 
concepts of adoption and inherited rank were among the dominant fea- 
tures of Hawaiian social customs and hero sagas. 

Hawaiians carried the concept of adoption to exceptional lengths as a 
means to establish and formalize new social relationships between 
members of the same social class, whether of commoners or of chiefs, in 
order to increase the number of individuals and families who looked to 
each other for emotional support and, in the most common form of 
adoption, economic support as well. For chiefs there was the further 
advantage of political and military support. The concept of adoption 
was used even to formalize close bonds of affection between two unre- 
lated and economically independent individuals. Two women might 
adopt each other as sisters, two men take each other as brothers, and a 
husband and wife become brother and sister. Further, a man or woman, 
boy or girl, regardless of age, might become the honorary, nonsexual 
spouse of a member of the opposite sex; or they might assume a role 
comparable to that of godparent or godchild to a member of the same or 
opposite sex. 

The Foster Child in Hawaiian Myths and Customs 37 

Additionally, an individual might adopt one or more of his contem- 
poraries as his "children," or an adult might become the adoptive parent 
of a younger adult; in each case, however, the adoptive parent trained 
and supported the adopted child and in turn was served by it. The con- 
cept of adoption was also carried into government, for a chief was met- 
aphorically the adoptive father of dependents on his lands and they 
were under his control. In various ways, then, fictive kinship bonds 
were created that were phrased in terms of adoption. 

The most prevalent pattern of adoption was that which under certain 
conditions could give both form and stimulus to a latent foster-child 
fantasy. This pattern involved parents giving up their children (male or 
female), albeit reluctantly at times, to be reared and economically sup- 
ported by their relatives or close friends of the same social class as them- 
selves, while they, in turn, adopted the children of others to rear. With 
rare exceptions, adopted children were not cut off from their natural 
parents, for the two sets of parents and their kin were in frequent con- 
tact with each other and the children. A royal child's adoptive parents 
who had charge of its personal and intimate daily care were supervised 
by the child's senior relatives to insure that the child was protected and 
trained for its future role. A child, male or female, who was a member 
of the chiefly class was taught both parents' genealogies and everything 
else such a child should know. A commoner's child learned about its rel- 
atives as well as the taboos and guardian gods it had inherited. 

Such in theory and usual practice was the pattern of adoption fol- 
lowed in adopting children at birth or soon thereafter. That a child 
knew the identity of both his adoptive and natural parents would not 
preclude his development in terms of the "family romance." In fact, a 
male would have two fathers to rebel against and two mothers whose 
affection he wished to reserve for himself. And there would still be the 
longing for the ideal father who was superior to both the natural and 
the adoptive one. 

Certain social conditions provided fertile ground in which a foster- 
child fantasy could grow. As has often been pointed out, the natural 
father's identity, unlike the mother's, was often uncertain. Some chil- 
dren were of double-headed paternity. Some children did not know the 
identity of their natural parents. This might happen if their adoptive 
parents took them to a distant locality or another island, where they lost 
contact with their natural parents and if neglected or mistreated ran 
away to become perhaps waifs adopted by strangers. Warfare, political 
intrigue, and ambition particularly endangered highborn children 
regardless of the care their adoptive parents and caretakers gave them. 

38 Pacific Studies, Vol. 10, No. 2— March 1987 

Records tell of such children being kidnapped, abducted, abandoned, 
possibly substituted for another, or completely lost from sight. 

And despite the attention the chiefly class gave to arranging mar- 
riages to maintain a pure line, misalliances and casual affairs cut across 
classes. Efforts to destroy the offspring at birth did not always succeed, 
especially if the mother was a commoner whose name the chief did not 
know. A survivor might or might not be able to prove his genealogical 
claim to rank and win his chiefly father's acceptance. The mother of a 
child "begotten by the roadside" would not know the father's identity or 

These social conditions indicate why in the Hawaiian sagas attention 
centers on only the father being unknown until he proves to be a ruling 
chief, a god, or important court retainer with the nominal title of chief. 
That the father-seeker is male except in one saga expresses the psychic 
superiority of men over women, who were regarded as inferior and pol- 
luting, regardless of rank. Nevertheless, that one saga suggests that the 
foster-child fantasy was also present among girls and women. 

What is known about the Hawaiian foster-child fantasy must be 
inferred from the sagas since there is no study of the fantasy among liv- 
ing Hawaiians. It is interesting, however, how frequently one hears an 
individual claiming, and perhaps with proof, to be a descendant of a 
renowned chief of the nineteenth century or even of three centuries ear- 
lier. Other chiefs besides Keawe and 'Umi scattered their favors so 
widely that the fantasy of exalted ancestry if not of parentage survives 
in modern society. What form the present fantasy of distinguished par- 
entage takes remains to be learned. 


1. Luomalal940, 1961b. 

2. Kirtley 1971:364, H1381. 2.2, H1381. 2.2.1, H1381. 

3. Kirtley 1971:202, D1602.2; 160, C939.3; Luomala 1955:161-178; Beckwith 1940: 
259-275; Beckwith 1930. 

4. Rank 1914:82. 

5. Lehrman 1927. 

6. This study took place in Oregon (U.S.). See Conklin 1920; Conklin 1935:140ff. 

7. See Luomala 1949, a monograph on Maui, and Luomala 1955:85-98 on Maui; 
Luomala 1955: 161-178 on Rata (cognate of Laka) . 

The Foster Child in Hawaiian Myths and Customs 39 

8. Beckwith (1940:483) was surely thinking of ni'aupi'o, a term for rank, when she 
wrote: 'The name Niauepo'o is a class title in Hawaii for chiefs of the highest rank, born 
from the marriage of close relatives among high chiefs." She spells the term for rank cor- 
rectly later (1951:13-14). Xi'aupi'o. literally, "bent coconut rib, i.e., of the same stock," 
identifies the "offspring of the marriage of a brother and sister, or half-brother and half- 
sister" (Pukui and Elbert 1957:245). Examples occur in the father-seeking narratives. 

9. See Beckwith (1940:478-488) on the stretching-tree shape-shifter and symbolism. 

10. Only versions with the father-seeker's query are listed below; each narrator tends to 
tell his version differently. Fornander (1880, vol. 2) has discussed 'Umi, Kiha, and Paka'a 
in their historical context. 

L'nii. Fornander 1917, vol. 4:178-235 (said to combine Malo's and Kamakau's 
accounts); Malo 1951:257-265; Pogue 1978:147-153; Kamakau 1961:1-21 (lacks the 
query, but a footnote, p. 1, cites other versions); Beckwith 1919:649-650; Beckwith 
1940:389-391, abstracts. 

Kiha. Thrum 1923:77-86, from The Polynesian 1 (1840); Thrum 1923:73-76, from 
Kuokoa, 18 Nov. 1865 (part of a version, no query); Beckwith 1919:650; Beckwith 
1940:387-389, abstracts; Kamakau 1961:22-33 (no query but a footnote, p. 22, cites other 
versions) . 

Pakaa. Thrum 1923:53-67; Rice 1923:69-89; Beckwith 1919:650-651; Beckwith 
1940:86-87, abstracts. 

Maui. Beckwith 1951:135-136, Beckwith translation; Beckwith 1940:227-229, from 
Ho'olapa; Luomala 1949:111-112, from Lili'uokalani. 

Laka. Thrum 1907:111-112; Beckwith 1930; Beckwith 1940:263-264, abstracts. 

Nakuemakapauikeahi. Rice 1923:19-31. 

KalanimanuVa. Fornander 1917, vol. 4:540-553; Beckwith 1919:657; Beckwith 
1940:479-480, abstracts. Kamakau (1961:169) mentions a place named Kalanimanui'a in 
'Ewa, Oahu, where a great battle was fought in 1794. 

NVauepo'o. Pukui 1933:179-185; Beckwith 1940:279, abstract. 

Laukiamanuikahiki. Fornander 1917, vol. 4:596-609; Beckwith 1919:655; Beckwith 
1940:513-514, abstracts. 

11. Dickey 1917:16-18; Beckwith 1940:231-232, abstract. 

12. Bastian 1883:278; brief reference, 232. 

13. Luomala 1961a: 155; Beckwith 1940:229. 

14. The two heroes are: 

Keaunini. Beckwith 1940:506-513, abstracts; Westervelt 1915:163-223; Thrum 1923: 
220-227; Fornander 1880, vol. 2:49, 56-57; Fornander 1919, vol. 6:345; M. K. Pukui, 
unpublished version. 

Namakaokapao'o. Fornander 1919, vol. 5:274-283; Beckwith 1940:480-481, abstract. 

15. Ku and Hina — with or without descriptive epithets relating to fertility of people, land, 
and sea — are a parental pair in religion, myth, and romance. Ku-waho-ilo (Maggot- 
mouthed Ku) is Kaanaelike's grandfather; she climbed to him on a stretching coconut tree 
and returned on his long tongue. His long tongue also scooped up the remains of 'Umi's 
half-brother, whom he sacrificed to the gods. On Ku and Hina, see Beckwith 1940:12-30; 
on floating islands and cloudlands, see ibid. , 67-80. 

40 Pacific Studies, Vol. 10, No. 2— March 1987 

16. A narrative without Naku'emaka concerns Anelike (Kaanaelike) and her earthly hus- 
band, a fisherman; after a separation she has him distinguish her from her eleven sisters 
(Green 1926:115-118). 

17. Burrows 1923:143-173, Fakaofo, Tokelaus; Smith 1913, vol. l:146ff., one of several 
Maori (New Zealand) versions. 

18. Mann 1937. 

19. Lowie 1930:460. 

20. Handy and Pukui 1958:71-72; Pukui, Haertig, and Lee 1972, vol. 2:36. See Howard 
et al. (1970) for a survey of traditional and modern customs of adoption and fosterage. 

21. Pukui, Haertig, and Lee 1972, vol. 1:49, 51-53; Kamakau 1964:26-27. 

22. Pukui and Elbert 1957:52. 

23. Pukui, Haertig, and Lee 1972, vol. 1:131-132. A legally adopted child is called keiki 
hanai ho'ohiki (Pukui and Elbert 1957:64). On luhi, see Beckwith 1919:595. 

24. Pukui, Haertig, and Lee 1972, vol. 1:131-132. 

25. Malo 1951:260, 265 n.4 by Emerson; Handy and Pukui 1958:68-69. Namakaoka- 
pao'o's father was also called makua kane kolea (Fornander 1919, vol. 5:277). 

26. Malo 1951:54-56, 60, 63 n.17; Fornander 1880, vol. 2:28-30; Kamakau 1964:4-6; 
Beckwith 1951:11-14. 

27. Fornander 1880, vol. 2:27-29. As Fornander said, "Once a chief always a chief," and 
no crime he committed could alter that fact for him and his children. On Kamehameha, 
see Kamakau 1961:208, 260; Fornander 1880, vol. 2:320. 

28. Kamakau 1961:188-189. 

29. Fornander 1880, vol. 2:260. 

30. Kamakau 1961:68. Kahekili was Kekaulike's son but the twins' father was said to have 
been Keawe-poepoe (Kamakau 1961:31; Fornander 1880, vol. 2:154 n. 4). 

31. Handy and Pukui 1958:54. 

32. Kamakau 1961:385. Kamakau points out the advantages of Kaheiheimalie's double- 
headed paternity and the kin she shared with Kamehameha. 

33. Ibid.:347. Fornander 1880, vol. 2:108, does not mention the double-headed paternity. 

34. Malo 1951:56. 

35. Ibid. 

36. Handy and Pukui 1958:79. 

37. Malo 1951:70; Handy and Pukui 1958:196. 'Umi, Liloa's favorite, was, Liloa said, 
"the boy that will make my bones live, this child of an owl" (Kamakau 1961:9). Liloa, 
calling him "owl's child," perhaps one of his heir Hakau's insults about Umi's mother's ori- 
gin, seems to mean that nonetheless 'Umi would one day do him more honor than 
highborn Hakau who was a pleasure-seeker, woman-chaser, and cruel to commoners. The 

The Foster Child in Hawaiian Myths and Customs 41 

expression "the bones live" also is used by elders about a considerate and kind child 
(Handy and Pukui 1958:179). 

38. Kamakau 1961:4. On kauwa, see Malo 1951:68-72; Handy and Pukui 1958:79, 204- 
205; Kamakau 1964:8-9. 

39. Fornander 1880, vol. 2:112-113. 

40. Kamakau 1961:152-153. 

41. Beckwith 1951:30. See Fornander (1880, vol. 2:139, n. 1) for an example of such a 
controversy between Kamehameha's supporters and his opponents. 

42. Wilkes 1856, vol. 4:31. Kamehameha's children by Ke'opuolani superseded his older 
children by wives of lesser rank. 

43. Kamakau 1961:4. 

44. Pukui and Elbert 1957:162; Malo 1951:55; Kamakau 1961:1, 3, 15, 242; Kamakau 
1964:4, 6; Fornander 1917, vol. 4:238 n.2. 

45. Kamakau 1961:3. 

46. Ibid.:19. 

47. Fornander 1917, vol. 4:238, chant and n.2; on Pi'ilani's mother, see Fornander 1880, 
vol. 2:83. 

48. Handy and Pukui 1958:168. 

49. Kamakau 1961:25-32; Thrum 1923:73-76. Accounts vary, but according to Fornan- 
der (1880, vol. 2:206), Kumaka was Kahu'akole's sister. 

50. Kamakau 1961:56. 

51. Ibid.:32. The unfortunate youth's name was Aihakoko. According to Fornander 
(1880, vol. 2:103-104), Aihakoko was 'Umi's daughter, not son, but met a "tragical end" 
on Maui. Thrum (1923:85-86) states that Kiha's counselors had him let 'Umi's two sons 
rule instead of himself during 'Umi's lifetime. 

52. Kamakau 1961:20; Fornander 1917, vol. 4:230-232 (from Kamakau). 

53. Kamakau 1961:347. 

54. Ibid.:391, 393, 394 (Kekauluohi's birth); 253, 341, 394, 395 (her mother Kaheiheima- 
lie and Kamehameha I); 279-280 (Kekauluohi's marriages, adopted chidren). 

55. Kamakau 1961:279-280, 290; Ii 1959:152. Kina'u herself had been reared by her 
maternal aunt, Peleuli, one of the wives of Kamehameha I (Kamakau 1961 :346) . 

56. Kamakau 1961:347. 

57. Ii 1959:15. 

58. Beckwith 1932:126, 163 n. 55; Kamakau 1961:259. 

59. Kamakau 1961:263-265, 269, 337-339; 348 (on Moses). 

60. Beckwith 1932:122, 123 (kahu alii and kahu hanai); Kamakau 1961:288, 304, 336- 
337 (Hoapili as guardian); 212, 215 (Hoapili as caretaker of Kamehameha's bones); 261, 

42 Pacific Studies, Vol. 10, No. 2— March 1987 

263 352, 393 (Hoapili as husband of Ke opuolani and Kaheiheimalie). Hoapili also helped 
rear Lot Kapualwa, later Kamehameha V. 

61. Kamakau 1961:260. Kaheiheimalie, it will be recalled, refused to let Kamehameha I 
adopt Kekauluohi. 

62. Ii 1959:161, 163, 164, 166, 167. Both Dr. G. P. Judd and Ii advised Klna'u not to take 
the infant to Kuakini on Hawaii. Kamehameha III later appointed Ii kahu of the royal 
and other highborn children, some from the outer islands, attending the missionary Royal 
School in Honolulu. Ii devotes most of chapter 12 (161-177) to his ward Victoria Kama- 
malu (1838-1866). 

63. Handy and Pukui 1958:48-49; Pukui, Haertig, and Lee 1972, vol. 1:192, 202-203. 

64. Kamakau 1961:347; Kamakau 1964:6, 10. 

65. Malo 1951:247-251; Fornander 1880, vol. 2:40-41, 74; Kamakau 1961:4. 

66. Fornander 1880, vol. 2:135-136; Kamakau 1961:67-69; Ii 1959:3, 6. 

67. Fornander 1880, vol. 2:134, 142-143; Ii 1959:3; Kamakau 1961:75. 

68. Ii 1959:3; Fornander 1880, vol. 2:143-144; Kamakau 1961:75-76. 

69. Fornander 1880, vol. 2:217-226; Kamakau 1961:128-138. 

70. Kamakau 1961:259. 

71. Ibid.:379-380. 

72. Ibid.:347-348. 

73. Kamakau 1964:26-27; Handy and Pukui 1958:71, 101-102. 

74. Green 1926:71-79. 

75. Handy and Pukui 1958:44, 71-72 (quotations); see also Pukui, Haertig, and Lee 1972, 
vol. 1:167; Pukui and Elbert 1957:115. Kenn (1939:48) differs from Handy and Pukui on 
ho'okama, saying that if a childless chief took a relative's child to rear with the intent of 
passing on his title, privileges, and possessions to it and the parents surrendered their 
rights, the process was called ho'okama, a term also used, he said, for modern legal adop- 
tion. Hanai to him implies a less complete transfer of parental rights. 

76. Kamakau 1961:6. For accounts of the adoption, see Fornander 1917, vol. 4:201, 203, 
221); Kamakau 1961:5-7. 

77. Fornander 1917, vol. 4:182 n. 2 by Thomas Thrum. 

78. Handy and Pukui 1958:73. They also discuss other categories of friendship. 

79. Fornander 1917, vol. 4:29-30. 

80. Handy and Pukui 1958:55. 

81. Handy and Pukui 1958:54-56, 57, 73; Pukui, Haertig, and Lee 1972, vol. 1:167; 
Pukui and Elbert 1957:327. 

82. Thrum 1923:73; Kamakau 1961:25; Pukui and Elbert 1957:108. 

The Foster Child in Hawaiian Myths and Customs 43 


Bastian, A. 1883. Inselgruppen in Oceanien: Reiseergebnisse und Studien. Berlin: 

Beckwith, M. W. 1919. The Hawaiian Romance of Laieikawai, by S. N. Haleole. Wash- 
ington D.C. : Bureau of American Ethnology Report 33. 

. 1930. "He Moolelo no Laka Wahieloa," from Jonah Kaiwaaea, with English 

translation of "The Story about Laka and Wahieloa." Notebook Vol. 5 (Ms.), in 
Bishop Museum Archives. 

. 1932. Kepelino's Traditions of Hawaii. B. P. Bishop Museum Bulletin 95. Hono- 
lulu: Bishop Museum. 

. 1940. Hawaiian Mythology. New Haven: Yale University Press. Reprint. Hono- 
lulu: University Press of Hawaii, 1970. 

. 1951. The Kumulipo. A Hawaiian Creation Chant. Chicago: The University of 

Chicago Press. Reprint. Honolulu: University Press of Hawaii, 1972. 

Burrows, W. 1923. "Some Notes and Legends of a South Sea Island. Fakaofo of the Toke- 
lau Group." Journal of the Polynesian Society 32: 143-173. 

Conklin, E. S. 1920. "The Foster-Child Fantasy." American Journal of Psychology 31:59- 

. 1935. Principles of Adolescent Psychology. New York: Henry Holt & Co. 

Dickey, L. A. 1917. "Stories of Wailua, Kauai." Hawaiian Historical Society 25th Annual 
Report, 14-36. 

Fornander, A. 1880. An Account of the Polynesian Race. Its Origins and Migrations and 
the Ancient History of the Hawaiian People. Vol. 2. London: Kegan Paul. Reprint. 3 
vols, in 1. Rutland, VT: Tuttle, 1973. 

. 1916-1920. Fornander Collection of Hawaiian Antiquities and Folk-lore. Edited 

by T. G. Thrum. B. P. Bishop Museum Memoirs 4-6 (vol. 4, 1917; vol. 6, 1919). 
Honolulu: Bishop Museum. 

Green, L. C. S. 1926. Folk-tales from Hawaii. Publications of the Foundation, second 
series, no. 7. Poughkeepsie, N.Y.: Vassar Folk-lore Foundation. 

Handy, E. S. C, and M. K. Pukui. 1958. The Polynesian Family System in Ka-'u, 
Hawaii. Wellington, N.Z.: The Polynesian Society. 

Howard, A., et al. 1970. "Traditional and Modern Adoption Patterns in Hawaii." In 
Adoption in Eastern Oceania, edited by V. Carroll. Association of Social Anthropol- 
ogy of Oceania Monograph 1:21-51. Honolulu. 

Ii, J. P. 1959. Fragments of Hawaiian History. Honolulu: Bishop Museum Press. 

44 Pacific Studies, Vol. 10, No. 2— March 1987 

Kamakau, S. M. 1961. Ruling Chiefs of Hawaii. Honolulu: Kamehameha Schools Press. 
1964. Ka Po'e Kahiko. The People of Old. B. P. Bishop Museum Special Publica- 

tion 51. Honolulu: Bishop Museum Press. 

Kenn, C. W. 1939. "Some Hawaiian Relationship Terms Re-examined." Social Process in 
Hawaii 5:46-50. Honolulu: Sociology Club of the University of Hawaii. 

Kirtley, B. F. 1971. A Motif-Index of Traditional Polynesian Narratives. Honolulu: Uni- 
versity of Hawaii Press. 

Lehrman, P. R. 1927. "The Fantasy of Not Belonging to One's Family." Archives of Neu- 
rology and Psychiatry 16:1015-1023. 

Lowie, R. H. 1930. "Adoption-Primitive." In Encyclopaedia of Social Sciences, edited by 
K. A. Seligman. New York: Macmillan. 

Luomala, K. 1940. "Notes on the Development of Polynesian Hero-Cycles." Journal of the 
Polynesian Society 49:367-374. Reprinted 1941 in Polynesian Anthropological Stud- 
ies, Polynesian Society Memoir 17. 

. 1949. Maui-of-a-Thousand-Tricks. His Oceanic and European Biographers. B. P. 

Bishop Museum Bulletin 198. Honolulu: Bishop Museum. 

. 1955. Voices on the Wind. Polynesian Myths and Chants. Honolulu: Bishop 

Museum Press. 

— . 1961a. "A Dynamic in Oceanic Maui Myths: Visual Illustrations with Reference 
to Hawaiian Localization." Fabula 4:137-162. 

— . 1961b. "The Fantasy of Distinguished Parentage in Polynesian Narratives" 

(abstract). In Abstracts of Symposium Papers, 10th Pacific Science Congress, p. 93. 

Malo, D. 1951. Hawaiian Antiquities (Moolelo Hawaii). Translated by N. B. Emerson, 
1890. B. P. Bishop Museum Special Publication 2. Second edition. Honolulu: Bishop 
Museum. First published in 1903, different pagination. 

Mann, T. 1937. Freud, Goethe, Wagner. New York: A. A. Knopf. 

Pogue, J. F. 1978. Moolelo of Ancient Hawaii. Translated by C. W. Kenn. Honolulu: 

Pukui, M. K. 1933. Hawaiian Folk Tales. Publications of the Foundation, third series, no. 
13. Poughkeepsie, N.Y.: Vassar Folk-lore Foundation. 

Pukui, M. K., and S. H. Elbert. 1957. Hawaiian- English Dictionary. Honolulu: Univer- 
sity of Hawaii Press. 

Pukui, M. K., E. W. Haertig, and C. A. Lee. 1972. Nana i ke Kumu (Look to the 
Source). 2 vols. Honolulu: Hui Hanai. 

Rank, O. 1914. The Myth of the Birth of the Hero. New York: Nervous and Mental Disease 

Rice, W. H. 1923. Hawaiian Legends. B. P. Bishop Museum Bulletin 3. Honolulu: Bishop 

The Foster Child in Hawaiian Myths and Customs 45 

Smith, S. P. 1913. Lore of the Wfwre Wananga. 2 vols. Polynesian Society Memoirs 3 and 
4. New Plymouth, N.Z. 

Thrum, T. G. 1907. Hawaiian Folk Tales. Chicago: A. C. McClurg. 

.1923. More Hawaiian Folk Tales. Chicago: A. C. McClurg. 

Westervelt, W. D. 1915. Legends of Gods and Ghosts. Boston: Geo. E. Ellis. 

Wilkes, C. 1856. Narrative of the United States Exploring Expedition. 5 vols. New York: 
G. P. Putnam. 


Keith L. Morton 
California State University, Northridge 

In 1951 Douglas Oliver wrote of Tonga: 

From the decks of a copra steamer pulling in to Nukualofa, this 
capital of the Kingdom of Tonga looks more like Cape Cod than 
South Seas. The illusion remains even after stepping ashore, 
because Tongans learned long ago that the easiest way to 
remain Tongan is to appear Western. (1961: 179) 

Oliver's observation rings with considerable truth even today, since 
Tonga and Tongans project a strong impression of conservatism and 
traditionalism. Despite Tonga's apparent continuity in life-style, the 
structure of Tongan society has been profoundly transformed by the 
processes of state formation, particularly the centralization and institu- 
tionalization of power and authority. George Marcus (1978) has de- 
scribed the consolidation of political power and the movement from tra- 
ditional chieftaincy to constitutional monarchy. These changes, at the 
societal level, have also transformed local organization by restructuring 
the social relations of production. 

Ethnographers working in Tonga during the last few decades have 
noted several features of Tongan local organization that suggest a strong 
trend toward social relations that are individualized, optative, and 
emphasize horizontal rather than vertical social relations. These obser- 
vations, presented as absences of expected patterns, have contrasted 

Pacific Studies, Vol. 10, No. 2— March 1987 


48 Pacific Studies, Vol. 10, No. 2— March 1987 

with what ethnographers understood to be traditional social relations. 
As early as the 1940s it was apparent that lineage or rammage organiza- 
tion was of little significance at the local level. The Beagleholes wrote 
that "there are no strong lineage-feelings among the villagers, nor any 
strong lineage- groupings. None of them had any association with, nor 
any interest in, the classic lineages of Tonga" (1941:71). Twenty-five 
years later, Aoyagi found a similar situation in another village: "an 
internal stratified system as the result of ramification . . . cannot be 
possibly found in this village" (1966:175). Kaeppler (1971) and Morton 
(1972) found that commoners were only weakly and occasionally affili- 
ated with titled persons and saw whatever association they had with 
nobles as a consequence of residence, rather than consanguinity. 

Perhaps the most detailed description of Tongan local organization is 
Decktor-Korn's dissertation, "To Please Oneself: Local Organization in 
the Tonga Islands" (1977). The author relies on the concept of "loose 
structure" to help depict the flexible and optative nature of local organi- 
zation. Decktor-Korn's material suggests that local organization consists 
of an amorphous association of households and larger, kindred-based 
residence units that are weakly defined, highly flexible with respect to 
both structure and membership, and function with a minimum of verti- 
cal integration. Beyond kinship are a number of voluntary associations 
and church-affiliated groups that exhibit even greater structural flexi- 
bility and less continuity in membership than do kin-based social 
groups. Decktor-Korn also notes that the village itself is not an impor- 
tant social unit, but rather an administrative unit within a highly cen- 
tralized and somewhat remote governmental structure. 

A Tongan scholar, 'Epeli Hau'ofa, observes and laments the extent of 
social change in Tonga. He describes the kinship system as "coming 
under heavy pressure . . . from overpopulation, increased monetiza- 
tion, and pressure on resources." According to Hau'ofa, the results are a 
movement toward nuclear family units rather than extended families 
and less interest in the affairs of kin beyond the household. His general 
conclusion is that Tonga is currently "in an era of uncertainty and con- 
fusion" (1978:16). 

There is, then, substantial agreement among ethnographers regard- 
ing the general features of modern local organization. These features 
include the weakened, if not absent, vertical social relations among 
social strata that were historically present (see Gifford 1929; Goldman 
1970; Sahlins 1958). Kin groups larger than nuclear-family households 
are weakly defined in both structure and function. Local organization is 
household centered and social relations beyond the household tend to be 

The Atomization of Tongan Society 49 

optative, temporary, and ad hoc. These characteristics are the emerging 
form of social relations at the local level and constitute a trend toward 
atomization. Rubel and Kupferer defined an atomistic society as "a 
society in which the nuclear family represents the major structural unit 
and indeed, almost the only formalized social entity." Also included in 
Rubel's and Kupferer's conception of the atomistic society are several 
antisocial psychological characteristics: "contentiousness," "suspicious- 
ness," and "invidiousness" (1968:189). These psychological traits are not 
generally characteristic of Tongan peasants, and the use here of the term 
"atomistic" is limited to its structural aspect. 

Atomization is a widespread, if not a worldwide, phenomenon that 
has usually been associated with industrialization. Weber (1950:111) 
suggested that both the size of the family and its cohesiveness were 
reduced by industrialization. Goode (1963:6) also notes that the nuclear 
family usually becomes independent of more inclusive social groups as 
industrialization expands. At the most general level, the atomization 
process in Tonga is possibly due to the incorporation and institutionali- 
zation of many aspects of Western culture, including Christianity, a 
Western educational system, and a political system that has greatly 
reduced the privileges of the traditional elite (Marcus 1978). Despite the 
potential influence of the variety of Western institutions that have been 
incorporated into Tongan society, I argue in this paper that the primary 
cause of atomization is the change in the social relations of production 
— specifically, the adoption of individualized land tenure. Individual- 
ized land tenure has established the economic conditions necessary for 
independent, nuclear-family households by functionally replacing tra- 
ditional social relations in the production process. 

The social implications of Tonga's individualized land-tenure system 
have been variously portrayed by authors dealing with the subject. 
Nayacakalou (1959) noted the predominance of nuclear-family house- 
holds, but did not connect this phenomenon with individualized land 
tenure. Instead, he interpreted modern land-tenure arrangements to be 
only a regularized and codified version of traditional land tenure. How- 
ever, this view can only be sustained by limiting the evidence to the 
immediate relationship between the commoner and the land; that is, in 
both traditional and modern circumstances individuals were ultimately 
assigned to work one piece of land. In a later study of land tenure, 
Maude noted the breakdown of the extended family and the role of 
modern economic circumstances in causing that process: "The main 
factors in this change in household form have been the change in the 
system of land tenure, the development of cash production, and a 

50 Pacific Studies, Vol. 10, No. 2— March 1987 

weakening in the ties which formerly bound the members of the 
extended family together" (1965:50). 

Joining the influence of the land-tenure system are population 
growth and internal migration, which result in village populations 
largely consisting of households with relatively few kin ties to one 
another. Villages themselves are relatively new features of Tongan social 
life and were originally established as fortresses in the first half of the 
nineteenth century, a period of internecine warfare. With the cessation 
of warfare after 1852, fortifications were no longer necessary, but the 
population remained clustered in villages (Kennedy 1958:163-165). 
Because villages were not part of traditional social organization and 
have few functions as social units in modern Tonga, their internal char- 
acteristics vary markedly in accord with local histories. 

In a previous article (Morton 1978) I described Tonga's economy as 
monetized but not commercialized. The limited development of a com- 
mercial economy is a significant constraint on the potential for econom- 
ically independent households and on the atomization process. Without 
the expectation of consistent or adequate income from participation in a 
commercial economy, Tongan peasant households rely on their social 
ties with other households to ameliorate temporary or long-term short- 
ages of subsistence goods. More generally, household interdependence 
results from the impossibility of long-term household survival by exclu- 
sive reliance on its own resources. Together, these conditions constrain 
and contradict the process of atomization by forcing interdependence 
and integration of households at the level of exchange. It is this phe- 
nomenon that I previously referred to as the communal economy (Mor- 
ton 1978). 

Tongan local organization, thus, consists of households that are 
potentially independent at the level of production but necessarily linked 
at the level of exchange by the impossibility of long-term economic 
independence. Furthermore, the social relations relied upon as links 
between households are not the social relations of traditional Tongan 
society, but social relations that conform to the modern circumstance of 
individualized production. The following analysis of production, kin- 
ship, and exchange is intended as further elucidation of this view of 
Tongan local organization. 

Tradition and Transformation 

The political transformation of Tonga began in the first half of the nine- 
teenth century with the ascendancy of Taufa'ahau to the position of Tu'i 

The Atomization of Tongan Society 51 

Kanokupolu, the most politically powerful of ancient Tu'i lines. By 
1875, with the aid of Wesleyan missionaries, Taufa'ahau was able to 
fully consolidate his political power, and in that year he became Tonga's 
first constitutional monarch. Taufa'ahau deprived most traditional 
chiefs of their power over commoners and sharply limited the power of 
those he retained as a hereditary nobility. After 1875 the nobility was 
economically dependent on the Crown rather than on their traditional 
authority over commoners, and commoners were emancipated from 
their traditional duties to chiefs. Marcus (1978) fully describes the nine- 
teenth-century events that led, in the twentieth century, to the restruc- 
turing of Tongan society. 

Prior to the political events of 1826-1875, Tongan society exhibited 
the general characteristics of Polynesian society: ranked social strata 
with genealogical position rationalizing differential access to political, 
economic, and religious power. While there is some ambiguity and dis- 
agreement concerning exact numbers of strata and the nature of rank 
within strata, the general structure of precontact Tongan society is rea- 
sonably evident and has been described by Gifford (1929), Goldman 
(1970), Sahlins (1958), and Mariner (Martin 1817). Uppermost in honor 
and sanctity was a line of paramounts, tu'i; a second stratum consisted 
of chiefs, 'eiki; and a third elite stratum, matdpule, acted as chiefs' assis- 
tants and attendants. The fourth major stratum, tua, or commoners, 
made up the bulk of the population. Gifford (1929) and Goldman 
(1970) give detailed accounts of the relationship among the elite strata. 
Here, it need only be said that Tonga's history from about 1450 to the 
modern postcontact period reveals a pattern of competition for power 
manifested in warfare, political assassination, and apparent manipula- 
tion of genealogical traditions in order to capture and consolidate politi- 
cal hegemony. The trend in this four-hundred-year period was a grad- 
ual reduction in the number of elite, titled lineages and the absorption 
of the power and property of weaker lines by more successful ones. 

The great division in Tongan society was between commoners and all 
chiefly strata. Every major aspect of Tongan culture manifested this 
division in qualitative and absolute terms. In religious thought the souls 
of all elite persons were immortal, leaving the body at death and exist- 
ing forever in Pulotu, the place of the dead. Commoners were thought 
to possess only a life force that deteriorated with the body after death 
(Mariner, in Martin 1817:100). Politics and warfare were strictly the 
prerogative of titled persons, with commoners acting as soldiers of the 
lowest rank. Materially, commoners were probably reasonably well-off, 
but they could not hold rights in objects of wealth and held no rights in 

52 Pacific Studies, Vol. 10, No. 2— March 1987 

resources beyond those assigned or allowed by chiefs. In the social 
sphere, commoners were not people of little honor, they were people of 
no honor. Commoners served chiefs as laborers and warriors, and lived 
as tenants on land controlled by their more socially esteemed and politi- 
cally powerful kin. 

Power flowed down the line of hierarchy, and in principle each 
superior down to head of household had great powers over the 
life and property of his subordinates. In practice, it was the 
commoner who was the universal subject of all powers, and it 
was the upper ranks who had the forcefulness to evade subordi- 
nation. In the system which had established the crude pattern 
of master and victim as a relationship akin to that between 
awesome god and a human being, the commoner was the vic- 
tim. (Goldman 1970:303). 

Drawing on the work of Rousseau (1978), Gailey (1981) has applied 
the concept of estate to Tongan stratification. "If the Tongan kin groups 
were stratified, but the strata were determined through kinship rela- 
tions common to both strata, they were not classes" (Gailey 1981:40). 
Estate systems are "a form of social stratification in which the strata are 
jurally defined, and where strata present a significant homology with 
the system of relations of production. The ideology of the system legiti- 
mizes inequality" (Rousseau 1978:87). The concept of estate predates 
the idea of class and is associated with an organic rather than a layered 
model of society (Fallers 1973:9). While separated by cultural differ- 
ences and hereditary status, estates are vertically integrated through 
shared rather than opposing ideologies. In Tonga, kinship provided the 
social linkage between estates and a shared ideology that supported sta- 
tus differences. 

At the local level, traditional Tongan society consisted of homesteads, 
'api, dispersed on the lands of a chief. Homesteads may have been occu- 
pied by patrilineally extended families (Maude 1965:50). Or, according 
to Latukefu (1967:3), the homestead consisted of a more loosely con- 
structed association of close kin, linked generationally through either 
patrilineal or matrilineal ties. In either case the social relations within 
the homestead and among closely related homesteads were structured 
by three principles of rank. Ego's patrilateral kin were 'eiki to him, that 
is, they outranked him and he owed them social respect. Matrilateral 
kin were of lower rank than ego and from them he could expect respect 
and material support. Within the sibling group, sisters outranked their 

The Atomization of Tongan Society 53 

brothers, and brothers owed both respect and material assistance to 
their sisters. Finally, there was a principle of seniority that ranked sib- 
lings of the same sex. These relations are described in greater detail by 
Latukefu (1967:4-7). 

I am more concerned here with the economic implications of rank- 
ing. Homesteads were under the authority of the 'ulu, head, who con- 
trolled labor within the homestead and oversaw the organization of 
production at the lowest level. Besides overseeing production on his 
homestead, the 'ulu was the social link between the homestead and the 
next higher level of organization, the fa'ahinga or matakali. The 
fa'ahinga consisted of a number of neighboring homesteads connected 
through common kinship. According to Latukefu (1967:8), "this was a 
named, exogamous group embracing several 'api headed by the 'ulumo- 
tua." The 'ulumotua, principal head, was probably from the chiefly 
estate and had to be at least of matdpule status. Like the 'ulu, the 'ulu- 
motua controlled production by distributing land among his home- 
steads, overseeing cultivation, and organizing labor. The 'ulumotua 
was either a close kinsman of a major chief, or the assistant, matdpule, 
of a major chief. In turn several faahinga, through common descent 
from the original title-holder, formed the kainga and were under the 
authority of the current title-holder, 'eiki. The kainga was the largest 
social unit associated with a discrete territory and the largest unit of 
local organization. It was the 'eiki of the kainga who received a grant of 
land from the Tu'i Tonga; an individual 'eiki might lose control of his 
land through the misfortunes of war or politics. Like the 'ulumotua and 
'ulu below him, the 'eiki took a direct interest in production, but at a 
more distant administrative level. For commoners, access to land 
depended on kinship links between the 'ulu at the lowest organizational 
level and the 'ulumotua and 'eiki at higher levels. 

While most of this construction is borrowed from Latukefu, I ques- 
tion his opinion that residence rather than kinship bound commoners to 
chiefs (1967:11). It is quite probable that most commoners did not know 
their genealogical connection to chiefs, but they did know their connec- 
tion to their 'ulu and their 'ulumotua. In a system of hierarchical rela- 
tions, it is only necessary for the participants to be aware of their rela- 
tions with those immediately above and immediately below them. Since 
the commoner was at the bottom of the genealogical scale, specific kin 
connections beyond the homestead and the fa'ahinga were not relevant 
to him. However, this lack of relevance for the commoner, participating 
at the lowest level of the society, does not imply that kinship did not pro- 
vide the overall structure of social relations. 

54 Pacific Studies, Vol. 10, No. 2— March 1987 

Production, distribution, and consumption were all organized by the 
rammage-like structure of traditional social relations. The territorial 
equivalent of the sociopolitical hierarchy was what Sahlins (1958:6-7) 
termed "overlapping stewardship." All land was first vested in the Tu'i 
Tonga, who granted tofia, estates, to principal chiefs, and they in turn 
granted use rights to their 'ulumotua; final allocation was made to 
heads of 'api by the 'ulumotua. The tenure rights of commoners were 
minimal since they could be evicted without recourse. In return for 
allowing access to his land, the chief demanded both goods and services 
from his tenants. A variety of goods were reserved exclusively for chiefly 
consumption, including certain types of yams, fish, turtles, and shell- 
fish. First fruits from all harvests were also for the use of chiefs. Goods 
produced at the local level moved up the hierarchy of statuses and sup- 
ported public works, warfare, craft production, and the lavish life-style 
of the chiefly estates. Labor for public works and for warfare was 
organized by the 'ulumotua and chiefs and either used locally or con- 
tributed for use at higher levels. 

The 'ulumotua exerted considerable influence over what crops were 
produced and what quantities were necessary for tribute and ceremo- 
nial offerings. Due to the abundance of resources and the general ease of 
production, it is doubtful that the material demands of the chiefs ever 
threatened the basic welfare of commoners. But commoners could not 
possess the prestige goods associated with chiefly status. Wealth was dis- 
tinguished from subsistence goods and consisted of ngatu (tapa cloth), 
fine mats, canoes, items of personal adornment, special architecture, 
and some food items, for example, pigs and certain yam varieties. 
Gailey (1981) maintains that kaloa, wealth, was produced exclusively 
by women, but this is doubtful since a variety of goods produced by 
male labor were also prestige goods, for example, yams, whales teeth, 
canoes, and houses. There was an emphasis on the circulation of goods 
as it relates to status and power, that is, status was manifest in the 
power to give and to demand goods. Thus the commoner was stigma- 
tized both in his role as provider and as receiver (Goldman 1970:301- 

Historically, Tonga's political economy was structured and domi- 
nated by the asymmetrical relationship between commoner and chiefly 
estates. Kinship served both to connect the estates materially and ideo- 
logically and to socially rationalize differential access to economic and 
political power. 

Of all the social and political changes that occurred in the establish- 
ment of Tonga's monarchy, the adoption of individualized land tenure 

The Atomization of Tongan Society 55 

was most significant in shaping modern local organization. As Marcus 
states, "It was the land arrangements, envisioned by the constitution, 
which were to give real substance to other clauses concerning individual 
rights and emancipation" (1978:516). Subsequent to 1875, land laws 
were enacted that granted usufruct directly to individual males, so that 
under the present land system every Tongan male, age sixteen and over, 
is entitled to 8V 4 acres of agricultural land, plus a small lot, 2 / 5 of an 
acre, on which to live within a village or town. This system is not com- 
pletely instituted and cannot be because of land scarcity (Maude 

Besides materially freeing commoners from their dependency on 
chiefs for access to land, the land laws have several significant social 
consequences. The modern land-tenure system not only frees the com- 
moner from the chief but also from the entire social hierarchy that pre- 
viously linked commoner and chief and defined social groups at the 
local level. The 'ulumotua no longer directs production, and a high 
proportion of households do not recognize any association with an 
ulumotua or a broader residence group. By emphasizing inheritance of 
land through primogeniture and allocating land only to males, the law 
undermines the traditional claim of women, as sisters, on the labor and 
resources of their brothers. It also depreciates the value of female links 
in establishing claims to land. Consequently, the land-tenure system 
exerts pressure on traditional forms of kin relations to transform them- 
selves into something approximating a Western model of kin relations, 
that is, households consisting of independent nuclear families domi- 
nated by males as husbands and fathers. 

Tonga's adopted land-use policy has also transformed social relations 
by clustering the population into villages. Prior to the modern period 
Tonga's population was dispersed in homesteads. It was only during 
periods of intense warfare in the nineteenth century that fortified vil- 
lages were temporarily established for protection. During the twentieth 
century villages became administrative units that could be overseen by 
a town officer, an elected official responsible to the central government. 
Internal migration and rapid population growth have created village 
communities in which kinship cannot function as the single, or even the 
dominant, mode of interaction. Mutual aid, cooperation, and other 
interactions generally are based as much on coresidence. 

Taken together, all aspects of Tonga's land-tenure system constitute an 
attempt to establish self-sufficient households, each with its own access 
to the means of production and each responsible for its own material 
welfare. This arrangement of the means of production and labor 

56 Pacific Studies, Vol. 10, No. 2— March 1987 

would, if achieved, approximate what Marshall Sahlins has termed the 
domestic mode of production, the D.M.P. However, the D.M.P. is only 
a logical possibility, an underlying structure; it cannot, in fact, be 

Clearly the domestic mode of production can only be a disar- 
ray lurking in the background, always present and never hap- 
pening. It never really happens that the household by itself 
manages the economy, for by itself the domestic stranglehold on 
production could only arrange for the expiration of society. 
Almost every family living solely by its own means, sooner or 
later, discovers it has no means to live. (Sahlins 1972:101) 

The long-term fragility of the D.M.P. constitutes a major constraint on 
the full development of economically independent households. In the 
case of Tonga there are other important constraints on the development 
of independent households. The lack of sufficient land to fully institute 
the system means that many households must rely on informal and 
extralegal methods to gain access to land. 

Tonga's marginal participation in the market economy also influences 
the nature of local organization and forces reliance on noncommercial 
forms of distribution and exchange. Several linked conditions have pre- 
vented the rapid development of a commercial economy in Tonga. Per- 
haps most significant is that land has not been commoditized; it cannot 
be legally transferred by sale. The nonexistent land market, combined 
with other government policies that discourage or prevent the intrusion 
of foreign capital, have largely prevented the establishment of large- 
scale enterprises. Commercial development has been limited to the ser- 
vice sector, retailing, and, to an even more limited extent, commercial 

During the period of this study (1970-1971), about 90 percent of 
Tonga's export income was derived from the sale of coconut products 
and bananas; the remainder was from the sale of other agricultural 
products and handicrafts. The 1966 census also reveals the predomi- 
nance of agricultural production in Tonga's economy. Sixty-seven per- 
cent of all adult males were engaged in agriculture while less than 3 per- 
cent worked in manufacturing and processing; services ranked as the 
second largest category of employment at just under 10 percent (Fiefia 
1968:28). Total exports for 1970 averaged about T$33 1 per person. 
These data reveal a commercial economy of small proportions domi- 
nated by agricultural production with minimal participation in com- 

The Atomization of Tongan Society 57 

mercial exchange. Obviously, Tongans do not rely on the production of 
cash crops or on the availability of wage labor to meet most of their 
material needs. Reproduction of the system at every level depends on 
subsistence needs being met through local production and distribution. 

Thus, as argued here, local organization in modern Tonga is more the 
result of accommodations to modern political and economic circum- 
stances than to conservation of traditional social relations. Maintenance 
of traditional social relations was not a viable twentieth-century alter- 
native because of changes in the political structure, particularly the 
reduction in chiefly authority. A more complete transition to a market 
economy and Western social relations was also not an option in the face 
of government policy that largely prevented investment by foreign capi- 
tal and limited development of commodity production within Tonga. 
The result has been a reformulation of local organization in terms of 
subsistence-oriented production carried on by household units disasso- 
ciated from traditional social relations. 

Local Organization and Production 

Despite their lack of easily discernible structure or function, Tongan vil- 
lages provide a pragmatic reference point to describe local organiza- 
tion. The village I selected for study, Matolu, 2 is located on the main 
island of Tongatapu. Matolu is distinctive in that the noble holding the 
estate on which the village is located resides in the village. All families 
in Matolu were members of the Wesleyan Methodist Church in the vil- 
lage. In most villages, religious affiliation is diverse and changes rapidly 
(Decktor-Korn 1977:170-195). 

Because of its proximity to Nuku'alofa and the fact that land is still 
available, Matolu has experienced very rapid population growth. 
Between 1956 and 1971 the population grew from 91 to 261, an increase 
of 287 percent. Mean household size grew from 4.8 to 6.2, and the num- 
ber of households increased from 19 to 42. This rate of growth is atypi- 
cal since Tonga's overall population grew by approximately 40 percent 
during the same period (Fiefia 1968:6). Only 7 of the 42 household 
heads were born in Matolu; 18 were born in other villages on Tonga- 
tapu, and the other 17 were born in villages located in other island 
groups. All of these men explained their change in residence by the 
availability of land and Matolu's proximity to schools and medical ser- 
vices. With the exception of one household, all of the "immigrant" fam- 
ilies have some attachment to earlier residents of the village, either 
through kinship or marriage. However, there was no pattern to the type 

58 Pacific Studies, Vol. 10, No. 2— March 1987 

of relationship utilized to establish residence by the immigrants. The 
one exception is the church steward, who is assigned to his position by 
the church administration. 

Matolu's rapid growth and the influx of immigrants resulted in some 
political and social tension. "Native" residents who can trace a long 
family association with the village are known as the "real Matolu peo- 
ple." More recent arrivals, even though they may have resided in the vil- 
lage for twenty years or more, are considered newcomers. This division 
is represented symbolically by the separation of the two groups in differ- 
ent graveyards. The division also appears in some local jokes, but there 
is no discernible effect on the quality of relationships between individu- 
als or households, or on the organization of the village itself. 

The availability of agricultural land is the main material concern of 
all families. First, they are concerned with immediate availability: Will 
they have land to use for subsistence gardening? Second is a long-term 
concern: Is land available for legal allotment? Other considerations are 
whether there are opportunities for wage labor, and whether there is 
ready access to health care and educational services. Because of the 
complexity of the decision-making process, it is difficult to classify resi- 
dence choices. With that in mind, 57 percent of the households were 
residing virilocally, 29 percent were neolocal, and 14 percent were 
uxorilocal. On the surface this distribution might be understood as con- 
sistent with traditional residence patterns. However, the primary con- 
cern is with access to land, and since transfer of land under the current 
land-tenure system gives strong preference to inheritance through pri- 
mogeniture, eldest sons would prefer to remain in close association with 
their father. 

After considering the origin of household heads, their residence 
choices, and their stated motives for selecting a particular village, it is 
apparent that kinship is playing only a limited role in determining vil- 
lage composition. Instead, access to land— which may be obtained in 
several ways, including use of kin relationships — is the primary concern 
in selecting the postmarital residence. All household heads in Matolu 
have access to gardening land, but only 13 hold legally registered allot- 
ments; 14 hold plots pending final approval by the noble, and 15 have 
temporary access to land through a kinsman or friend. Like most 
nobles, the noble holding the Matolu estate has been very slow to 
approve final allotment of land. Reluctance in this matter provides the 
noble with some control over peasant families vying for legal access to 
land. This manipulation by the noble is extralegal, but it is grudgingly 
accepted as his traditional right. 

The Atomization of Tongan Society 59 

Villagers express a preference for living in nuclear-family households, 
and of the 42 households in Matolu, 41 consist of nuclear families. Eight 
of these households include an adult, unmarried relative of either the 
household head or his wife. One household includes three conjugal 
pairs, but two of the couples are only recently married and, although 
intending to establish separate households, have not yet done so. 

All households engage in some cash-producing activities. These 
activities vary over time within the same household and between house- 
holds. Cash-crop production, mostly coconuts and bananas, fishing, 
production of handicrafts, and wage labor are the main sources of 
income. No households engage in any of these activities continuously; 
income-producing activities are always linked to specific targets of a 
nonsubsistence nature. Decktor-Korn (1977) found the same orientation 
to production for exchange. This pattern of intermittent involvement in 
the market economy can be explained by the small amount of income 
generated relative to time input plus the intermittent nature of income 
sources themselves: wage-labor jobs are likely to be temporary and the 
markets for agricultural products are sporadic. 

Of the 42 household heads in Matolu, 25 identified themselves as 
farmers, 6 as wage laborers, and 5 as fishermen; 2 others were working 
in New Zealand as wage laborers. The remaining 4 household heads 
consisted of minor church and government officials. This information 
suggests more specialization than actually exists, since all household 
heads, with the exception of the two absent from the village, gardened 
and relied on gardening to meet household subsistence needs. Strict 
occupational specialization in the market economy is impractical for 
most because most income sources are temporary; this is true of both 
wage-labor jobs and the production of cash crops. 

Production is carried out by households operating independently of 
one another. Household heads make their own decisions on such matters 
as the use of horticultural land, crop selection and rotation, and 
whether to plant cash or subsistence crops. Their decisions are not sub- 
ject to review or control by any higher social or political authority. For 
farmers who do not have their own allotments and rely on kin or nobles 
for access to land, this statement must be modified. These farmers are 
less likely to grow cash crops or to invest as much labor in the land as 
farmers working their own allotments, because their rights to the land 
are temporary. 

Labor is largely supplied by the household head working alone, occa- 
sionally assisted by other males in his household — sons or visiting kins- 
men. More rarely, wives will assist their husbands. Labor is often 

60 Pacific Studies, Vol. 10, No. 2— March 1987 

pooled for tasks that require large labor input in short periods, or for 
tasks that are extensive and considered boring. Most formally, labor 
pooling is organized as a kautaha toungaue, an association in which the 
participants agree to work for one another, as a group, in rotation. One 
individual is selected as timekeeper to help enforce the principle of 
equal contribution. During 1971 Matolu had two such groups, one hav- 
ing a membership of ten and the other a membership of seven. The tasks 
undertaken as a group were planting and hoeing. On the days the asso- 
ciation was to work, each member had one hour in the morning to work 
on his own garden plot; the remainder of the day the entire membership 
worked on the garden of one member. The order of rotation was estab- 
lished by the proximity of each garden to the village, group work begin- 
ning at the closest garden and ending at the most distant. 

Kinship was not a criterion for membership; instead the groups were 
formed on the basis of residence. All members of one group resided in 
the northern half of the village, and the members of the other group 
resided in the southern half. It is also important to note that these 
groups functioned without a leader. The timekeeper was selected 
because he was thought to be honest and sufficiently literate to keep 
records. Less than half the men chose to belong to a formal gardening 
association. However, there are less formal, smaller groups of friends 
and neighbors that work together. In these cases, the labor input of each 
should balance out but is not strictly accounted since companionship is 
the dominant concern. 

In contrast to traditional circumstances, kin groups, or even dyadic 
relationships with kin outside the households, play virtually no role in 
organizing production. The only kin group larger than the household is 
the matakali, and it is both corporately weak and structurally ambigu- 
ous. In the village of Matolu there are four recognized matakali. Of the 
42 households in the village, only 21 claim affiliation with a matakali. 
In each household claiming affiliation, either the household head or his 
wife is a consanguine of the 'ulumotua. The position of the 'ulumotua 
is not strictly determined by kinship or descent. Instead, the position 
seems to result from the merging of a consideration of village history 
and current political realities. Recognized 'ulumotua claim relatively 
long ancestral associations with the village and have some claim to sta- 
tus through seniority, title, or control of some official position such as 
town officer. These men tend to be the foci of status rivalry within the 
village, not because they are 'ulumotua, but by virtue of holding other 
political or social positions. Economically, they are sometimes signifi- 
cant in mobilizing goods and labor for ceremonial occasions and/or 

The Atomization of Tongan Society 61 

feasting, but they do not have the authority or personal power to do 
more than request contributions and cooperation. 

Historically both the matakali and the 'ulumotu'a were more socially 
significant and were apparently critical in organizing production at the 
local level. Latukefu (1967:8) states that in the past only titled persons 
could hold the position of 'ulumotu'a; this suggests that historically 
matakali were segments of ramages, ha' a. The 'ulumotu'a of precontact 
Tonga were quite powerful since they held the final rights to land in the 
system of overlapping stewardship. Under that system, the land of the 
matakali was distributed and cultivated under the supervision of the 

The modern system of land tenure has altered the chain of steward- 
ship so that the 'ulumotu'a no longer receives land to administer and no 
longer supervises production. The modern matakali, to the extent that it 
exists, has weakened and changed in both structure and function. Gif- 
ford (1929) and Latukefu (1967) both describe matakali as segments of 
larger lineal units, ha' a. Today the matakali are not segments of larger 
units, nor do they function directly in the production process. Also, the 
power of the 'ulumotu'a over members of the matakali has been signifi- 
cantly reduced, since they no longer rely on the 'ulumotu'a for access to 

In modern Tonga kinship has a limited and peripheral role in organiz- 
ing production, particularly in organizing labor and allocating access to 
land, the primary means of production. Consequently, its role in estab- 
lishing and defining local organization has also been significantly 
reduced. Kinship groups larger than the nuclear family and the house- 
hold are poorly defined and functionally ambiguous. Villages them- 
selves lack structure related to kinship. None of the foregoing is meant 
to claim that kinship is not an important aspect of modern Tongan social 
life; it clearly is. However, in removing traditional kin relations from 
production their material constraints on the community and the indi- 
vidual have been altered. 

Local Organization and Exchange 

Tongan local organization roughly approximates the household type of 
peasant political-economy described by Halperin (1977:291). In this 
type of peasant system, households are relatively autonomous at the 
local level, political power is a function of the traditional sociocultural 
system, and the local elite exert economic controls through taxation and 
authority over the distributive system. In the two other types of peasant 

62 Pacific Studies, Vol. 10, No. 2— March 1987 

systems defined by Halperin, administered community and commercial 
plantations, elites exert much more direct control over production itself. 
More significantly for this analysis, in a household-type peasant econ- 
omy the reproduction of the local community is left to the local commu- 
nity. This is the case in Tonga where the state provides little economic 
protection. Tongan households are left to their own social devices to 
protect themselves from short-term or long-term economic difficulties. 
The disadvantages of household independence are apparent to Tongan 
peasants. Illness, family discord, separation, and/or old age may 
seriously disrupt production and threaten the ability of the family to 
provision itself. Besides obvious material circumstances that slow or 
stop household production, there are a variety of social activities and 
religious duties that also disrupt production for varying amounts of 

The long-term unreliability of the domestic mode of production, 
combined with the limited development of a commercial economy, 
necessitates and produces the communal economy. The communal 
economy is not a total economy; it does not have its own form of pro- 
duction. Goods that enter the communal economy are goods produced 
for use by households; they are not produced specifically for exchange. 
The communal economy is a sphere of exchange in which goods move 
from household to household through the social linkages of kinship, 
friendship, and neighborliness. 

The ideology of the communal economy contrasts with the individu- 
alistic nature of production. Its widest principle is fetakoniaki, the 
spirit and reality of cooperation. More specifically, this ideology pre- 
vails upon individuals to materially assist kin, neighbors, and friends, 
particularly those who need assistance. It also calls upon the able indi- 
vidual to assist others who are socially close before using resources in 
other ways— for example, sharing food surpluses with neighbors rather 
than selling them. Tongan attitudes toward food are perhaps the best 
example of this ideology. Tongans view food as almost a free good. No 
one should be deprived of food; "come and eat" is probably the most 
frequent greeting in Tonga and is extended to strangers as well as close 
kin and friends. There is also a pervasive attitude that the selling of food 
is a breach in custom and borders on immorality. 

To determine the characteristics of Tonga's communal economy, I col- 
lected data on 604 transactions. The data were obtained from 40 house- 
holds in Matolu over a ten- week period. The sampling procedure was to 
visit each household at two- week intervals and elicit information on the 
last four transactions in which the household had been involved. Dupli- 

The Atomization of Tongan Society 63 

cate reports of the same transaction were removed from the sample. 
Collecting information from both principals involved in a transaction 
provided an opportunity to evaluate the validity of the data. The few 
discrepancies I found were minor and were probably due to honest dif- 
ferences of opinion or memory. There was some difficulty in collecting 
this information since Tongans considered it uninteresting to discuss the 
details of everyday exchange. Their interest focused on those transac- 
tions they considered socially significant. It was easier for an informant 
to remember the details of his gift of a large pig for his brother's wed- 
ding feast several years earlier than the basket of taro he gave to his 
neighbor two days before. By focusing my informants' attention on one- 
and two-day periods preceding the interviews, I obtained a sample 
more closely reflecting the general characteristics of the communal 
economy. Reliance on informants' selection of transactions would 
undoubtedly have skewed the sample toward transactions involving 
prestige items in a ceremonial context. In this sample only 7 percent of 
the cases were associated with a ceremonial context. Social contexts that 
do require gifts are life-crisis events — births, certain birthdays, wed- 
dings, and funerals. Guests attending feasts also provide their hosts with 
gifts of prestige foods or traditional craft items, particularly tapa and 
mats. However, the emphasis here is on more mundane transactions and 
their role in the communal economy. The great preponderance of these 
transactions, 93 percent, occurred in the course of daily patterns of 
interaction rather than during extraordinary ceremonial events. 

Virtually all of these transactions would be classified as generalized 
reciprocity (Sahlins 1965:147). Their two-sided quality is not apparent 
in a single transaction. At the moment of transfer there is, of course, a 
giver and a receiver, but over time this distinction is dissolved. The com- 
munity's expectation is that individuals will be involved in the commu- 
nal economy as both givers and receivers. There is some prestige 
attached to being a giver, and individuals who seem to consistently be 
receiving without good reason, such as physical disability, are the sub- 
ject of some gossip and may be subjected to mild public ridicule through 

Initiation of transactions is often subtle because "gifts" may be elic- 
ited by admiring a desired object or by mentioning a need. With this in 
mind, 26 percent of the transactions were initiated by a request and the 
remaining 74 percent were initiated by the giver. 

The types of material exchanged further suggest the importance of 
the communal economy in the day-to-day provisioning of households. 
Foodstuffs are the largest category and constitute 63 percent of the sam- 

64 Pacific Studies, Vol. 10, No. 2— March 1987 

pie. Nearly all of the foods exchanged were produced at the local level. 
Horticultural products made up 37 percent of the sample, fish and meat 
totaled 25 percent. Since there is no refrigeration in the village and no 
other preservation methods are used, meat and fish have to be con- 
sumed soon after they are obtained. Consequently, they are overrepre- 
sented in the sample relative to their actual frequency in the typical vil- 
lager's diet. In many households meat and/or fish is only consumed once 
or twice a week. 

The separation of the communal and commercial economies is mani- 
fest in the types of goods exchanged, the form of transactions, and their 
distinctiveness in social function. A variety of imported commodities 
have become necessities in household consumption patterns, for exam- 
ple kerosene, matches, salt, spices, cloth, and a variety of household 
utensils. These goods are treated differently than locally produced 
goods within the communal economy. Only 20 percent of the goods 
transferred were imported and, if tobacco products are excluded, only 
10 percent of the goods exchanged originated in the commercial sector. 
These commodities are usually given in very small quantities and are 
intended to tide the receiver over until supplies can be replenished 
through purchase. Expensive commodities such as radios, bicycles, and 
fine dishware are loaned with considerable frequency, but are rarely 
given freely. 

Money was exchanged in 11 percent of the transactions. These were 
small quantities, generally less than T$l. A request for money was 
always accompanied with a statement regarding its intended use. While 
difficult to quantify, exchanges involving money seemed to be treated 
with more specificity regarding use and conditions of repayment than 
did transactions involving goods. However, no transactions by sale were 
observed between co- villagers except those that occurred at the one vil- 
lage store. 

Production is organized and accomplished by households working 
independently. I argue here that this model of production as social pol- 
icy and as a system for allocating the means of production— land— has 
an atomistic effect on Tongan social life. In particular, it has to a large 
extent dissolved the material basis for hierarchical social relations. 
However, the inherent impossibility of production and reproduction 
based strictly on autonomous households necessitates socioeconomic ties 
among households. This necessity is made more pronounced by the lim- 
ited development of the commercial economy and the similarly limited 
development of social welfare policy and institutions at the society 
level. Reproduction of the community and the households themselves 
depends on a vigorous communal economy. 

The Atomization of Tongan Society 65 

Table i Transactions according to Kinship and Geographic Distance 





Exchange parties reside 
in same village 

150 (a) 

236 (b) 


Exchange parties reside 
in different villages 

160 (c) 







Note: Aa (kinship is dependent variable) = .29, Aa (location is depen- 
dent variable) = .05, A = .17, Chi-square = 66.48 (p <.001). 

Kinship and coresidence, together, are the social relations that struc- 
ture the communal economy. Table 1 categorizes transactions as to 
whether the principals were kin and whether they were coresidents. 
Lambda a (Aa), 3 Guttman's coefficient of predictability (Freeman 
1965:71-78), was used to examine the relationship between proximity 
and kinship. When Aa is calculated with kinship as the dependent varia- 
ble (i.e., kinship status is guessed with knowledge of residence), 29 per- 
cent of guessing errors are eliminated; when location is guessed with 
knowledge of kinship status, Aa is only 5 percent. This implies that 
shared kinship overrides distance and that coresidence overrides lack of 
kinship (compare cells b and c). The low frequency in cell d, less than 
10 percent of the sample, further confirms the significance of both co- 
residence and kinship as important determinants of exchange. 

Obviously, close proximity is likely to promote relationships of coop- 
eration and mutual assistance, and to a large extent villagers expect to 
rely on neighbors for material assistance. Interdependence of neighbor- 
ing households is symbolically recognized in the distribution of Sunday 
feast foods. Food for Sunday is usually prepared the day before since 
cooking and other forms of work are illegal on Sundays. Households 
prepare large quantities of one or two dishes and then a child is sent to 
deliver portions to several neighboring households. At the practical 
level this allows each household to have a variety of foods without the 
extra effort of preparing several different dishes. 

Fifty-one percent of the sample consisted of exchanges between kin, 
so in general it can safely be concluded that kinship is playing an active 
role in determining with whom one exchanges. The strongest tendency 
revealed in table 1 is that when exchange occurs between principals 
residing in different villages they are likely to be kin. Within the village 

66 Pacific Studies, Vol. 10, No. 2— March 1987 

kinship plays a much stronger role than the data suggest because the 
non-kin to kin ratio is quite high. Village demographics are such that 
relatively few kin are available within the village. 

The proportion of transactions that involve kin, however, only tell us 
that kinship is a significant feature in the communal economy and local 
organization. The exact kin relations involved and their relative propor- 
tions are much more revealing about the material role of kinship. Fifty 
percent of the exchanges occurring among kin are between parents and 
children or between siblings, that is, persons that once resided together 
in the same household and constituted a nuclear family. The frequency 
of exchange falls off dramatically with increasing degrees of consan- 
guinity. Overall, the pattern of exchange among kin is one in which the 
household organized around nuclear families emerges as the socially sig- 
nificant unit. Exchange within the communal economy both contra- 
dicts and reaffirms the atomistic influence of the organization of pro- 
duction. While it links households together in recognition of the 
interdependency of households, it is itself organized in terms of house- 
holds rather than more inclusive kin groups. 

Tongans have a strong conscious model of exchange that for them 
should be both prescriptive and descriptive. The model is based on kin- 
ship rank within the bilateral kindred, kainga, and specifies the direc- 
tion of the flow of goods, that is, goods should move from persons of 
lesser rank to persons of greater rank. Relative rank within the bilateral 
kindred is established by three criteria applied in the following order: 
agnates are of higher rank than ego and uterines are of lower rank; 
females occupying the same genealogical position as males are of higher 
rank than their male counterparts; and age outranks youth. So the 
eldest female in a sibling group outranks all of her siblings. In the 
parental generation father's sister is of particularly high rank because of 
her position as both agnate and female. Mother's brother, because of his 
position as both uterine and male, is the epitome of low rank within the 
kindred. Traditionally, the relationship between mother's brother and 
sister's son was known as the fahu relationship, meaning "above the 
law." In this relationship ego was allowed to exact unlimited goods and 
services from mother's brother. 

This model of exchange is understood by virtually all adult Tongans 
and is thought to characterize exchange behavior. It is discussed by 
E. W. Gifford (1929:17-19) and by Ernest and Pearl Beaglehole 
(1941:76-78), and finally by Adrienne Kaeppler. After describing kin- 
ship rank, Kaeppler states, "The economic implications of status within 
the system of exchanges among the Kainga can be characterized as indi- 

The Atomization of Tongan Society 67 

rect reciprocity, i.e. goods and services go from ego and his siblings to 
his patrilateral relatives, while he exacts goods and services from his 
matrilateral relatives" (1971:179). Whether Kaeppler's statement is cor- 
rect or not is not really at issue. It is certainly correct in that it repre- 
sents the way Tongans themselves view exchange between kin. How- 
ever, the actual pattern of transactions in my sample contradicts this 
view because, first, exchange with kin outside the first degree of collat- 
erally is relatively infrequent and, second, because there is not a one- 
way flow of goods within kin dyads involving differences in rank. The 
actual direction in which goods moved only conformed to the ideology 
of kin ranking in 37 percent of the transactions. Clearly, kinship rank is 
not significantly influencing the direction in which goods move within 
the communal economy. 

The ideology of kin rank is, of course, associated with traditional 
relations of production in which access to the means of production was 
defined by kinship status. Under modern circumstances it contradicts 
the actual organization of production and the atomistic character of 
local organization. Its survival and Tongans' belief in its efficacy possi- 
bly lies in its ideological-political function rather than its economic role. 
Tonga's modern government, although quite different from the tradi- 
tional political system, is largely rationalized in terms of traditional kin 
ranking and the associated system of stratification. "Former relations of 
production and other social relations do not disappear suddenly from 
history, but they are changed; they influence the forms and places 
which will assume and manifest the effects of the new conditions in 
material life, within the former social structure" (Godelier 1977:5). 

The shift in the economic functions of kinship are critical to under- 
standing the direction of social change in Tonga. All forms of kin rela- 
tions, beyond those within the nuclear family, have been disassociated 
from production, and it is this disassociation that produces the atomistic 
character of local organization. To a large extent kinship maintains its 
integrative function by facilitating mutual assistance between economi- 
cally interdependent households, but this is, at base, the limited kinship 
of the nuclear family. The social character of exchange within the com- 
munal economy disguises significant structural change in local organi- 


That traditional forms of social relations are transformed by the intru- 
sion of Western social models and a capitalist market economy is already 

68 Pacific Studies, Vol. 10, No. 2— March 1987 

well established in social science. The goal of this paper has been to 
more closely examine the dynamic relations among production, social 
organization, and exchange in the reorganization of Tongan local orga- 
nization. At the broadest level it can be concluded that local organiza- 
tion little resembles its historic and traditional counterpart. By individ- 
ualizing production, Tonga's modern land-tenure system has been 
central in weakening and dissolving not only relations between tradi- 
tional strata, but all relations that previously linked individuals into 
hierarchically organized, corporate, kin groups. Internal migration, 
stimulated by the search for land under the modern system, plus popu- 
lation growth, have also contributed as atomizing influences. 

Completion of the atomizing process has been forestalled by the eco- 
nomic impossibility of a domestic mode of production and by the lim- 
ited development of the commercial economy. The presence and vigor 
of a communal economy attests to the economic interdependence of 
households and provides data that demonstrate the importance of social 
relations based on coresidence and kinship defined in terms of nuclear 
families. Because exchange within the communal economy is facilitated 
by social relations rather than by commercial motives, it appears as a 
traditional feature of local organization. However, these are not the 
same social relations that traditionally organized production and ex- 

Local organization in Tonga is the result of the specific way in which 
Tongan society responded to its domination by Western culture and cap- 
italist economic relations. In Tonga's case indigenous political control 
was maintained, but only by adopting Western social and political 
models. Most relevant to the argument advanced here is the adoption of 
the concept of a freeholding peasantry tied directly to a centralized state 
bureaucracy. As proposed by Lingenfelter (1977:114), "The superordi- 
nate variable in change is the domination of the colonial power which 
restructures the indigenous societies to extract from them a surplus, 
which is politically defined, and idiosyncratic to each historical time 
and place." Despite the enigmatic character of colonialism, an examina- 
tion of the way in which colonialism reorganized indigenous labor and 
consequently the way in which it impinged on traditional social rela- 
tions provides some comparison. In Tonga labor was removed from the 
control of traditional social relations by providing individual males 
with legal rights in the means of production through the authority of a 
state. Despite the continued emphasis on production for use instead of 
production for exchange, traditional social relations have largely been 

The Atomization of Tongan Society 69 

Finney's analysis of socioeconomic change in Tahiti (1973) suggests 
that local organization in Tahiti and Tonga are broadly similar in the 
weakening or destruction of traditional social relations and their 
atomistic character in the modern period. In his examination of food 
exchange in Tahiti, Finney found that 80 pecent of these exchanges 
occurred between siblings or between parents and children. In Tonga, 
the comparable figure is 50 percent. Social change in Tahiti has, of 
course, been more dramatic and thorough than in Tonga, but this gen- 
eral state of affairs does not account directly for the atomistic character 
of Tahitian local organization. Tahitians have experienced a more direct 
intrusion of capitalism than Tongans by the thorough commercializa- 
tion of their internal economy. Finney's Tahitian peasants are oriented 
primarily to the production of cash crops and have attempted to indi- 
vidualize production themselves since the rewards of commercial pro- 
duction are distributed in terms of individual effort. Tahitian proletar- 
ians have lost their rights in the means of production and sell their labor 
directly in the labor market. Permanent social relations beyond the 
nuclear family become materially irrelevant where the commercial 
rewards for labor are continuous and meet subsistence needs. Unlike 
Tonga, Tahitian social organization is atomized by direct participation 
in a commercial economy, that is, subsistence needs are met by selling 
products or labor. 

Samoa, in contrast to both Tonga and Tahiti, is known for its cultural 
conservatism and social stability. Holmes (1971:101-104) notes a trend 
toward individualized land use. Increasingly, untitled men are allowed 
to use sections of family-held lands, move inland beyond the control of 
matai, and rely on wage labor. Matai are also finding it increasingly dif- 
ficult to control family lands. Despite these trends, 98 percent of the 
land in American Samoa, and 86 percent in Western Samoa, remains 
under the control of the matai (ibid.: 99). As concluded by Holmes 
(ibid.: 103), "land and social organization remain closely linked." Thus, 
in Samoa traditional social relations have remained central to produc- 
tion and continue to dominate local organization. 

The strength of traditional social relations in Samoa and their weak- 
ness in Tonga lends support to the hypothesis that the social relations of 
production, whether stable as in Samoa, or transformed as in Tonga, 
are central to the process of social atomization. Samoa has experienced 
more commercialization and industrialization than Tonga. It has also 
been exposed to the influences of Christianity, Western education, and 
tourism, yet has experienced less change at the level of local organiza- 
tion. The contrast between Samoan and Tongan local organization is 

70 Pacific Studies, Vol. 10, No. 2— March 1987 

starkly revealed in a comparison of extra-household local organization 
and its role in production. In Samoa, kin groups, 'ainga, rather than 
individuals hold land, and control of production is vested first in the 
matai of the 'ainga and finally in the village fono that establishes pro- 
duction goals for the matai(s). Further, the matai directly organizes 
labor within his 'ainga (Lockwood 1971:32-33). In Samoa, the village is 
a significant, if not central, level of organization in the production pro- 
cess. Through its hierarchy of kin relations the Samoan village organizes 
production by allocating land and labor and monitoring consumption. 
The Tongan village lacks any function directly related to production 
and, other than its passive role in establishing association through prox- 
imity, contributes nothing to the organization of labor. 

Tongan local organization is not a chaotic mixture of traditional and 
Western culture nor is it the result of Tongan attempts to remain tradi- 
tional; rather it is a result of the specific way in which Tonga has 
responded to its peripheral status in broader political and economic sys- 


I would like to thank Antonio Gilman and Gregory Truex for commenting on early drafts 
of this paper and discussing with me a variety of issues relating to the paper. Field research 
in Tonga was conducted in an eleven-month period in 1970-1971. It was sponsored by a 
National Institutes of Health Traineeship administered by H. G. Barnett, University of 

1. The basic unit of Tongan currency is the paanga, dollar, which was valued at T$1.00 
= U.S. $1.14 during this study. 

2. Matolu is a pseudonym for a village located on Tongatapu. 

3. Lambda a (Aa) is a measure of accuracy in guessing the value of one variable from 
knowledge of the value of another variable. Thus Aa reflects asymmetrical associations. 
Lambda (A) reflects symmetrical associations, that is, the result of guessing both ways. 


Aoyagi, Machiko 

1966 "Kinship Organization and Behavior in a Contemporary Tongan Village.'' Jour- 
nal of the Polynesian Society 75:141-176. 

Beaglehole, Ernest, and Pearl Beaglehole 

1941 Pangai: Village in Tonga. Wellington: The Polynesian Society. 

The Atomization of Tongan Society 71 

Decktor-Korn, Shulamit R. 

1977 "To Please Oneself: Local Organization in the Tonga Islands." Ph.D. dissertation, 
Anthropology Department, Washington University (St. Louis, Mo.). 

Fallers, Lloyd A. 

1973 Inequality: Social Stratification Reconsidered. Chicago: University of Chicago 

Fiefia, SioneN. 

1968 Report on the Results of the 1966 Census, Kingdom of Tonga. Nuku'alofa: Gov- 
ernment Printing Office. 

Finney, Ben R. 

1975 Polynesian Peasants and Proletarians. Cambridge, Mass.: Schenkman. 

Freeman, Linton C. 

1965 Elementary Applied Statistics: For Students in Rehavioral Science. New York: 

Gailey, Christine Ward 

1981 " 'Our History is Written in Our Mats': State Formation and the Status of 
Women in Tonga." Ph.D. dissertation, Anthropology Department, New School 
for Social Research (New York, N.Y.). 

Gifford,E. W. 

1929 Tongan Society. Bishop Museum Bulletin No. 61. Honolulu: Bishop Museum 

Godelier, Maurice 

1977 Perspectives in Marxist Anthropology. Cambridge: Cambridge University Press. 

Goldman, Irving 

1970 Ancient Polynesian Society. Chicago: University of Chicago Press. 

Goode, William 

1963 World Revolution and Family Patterns. New York: Free Press. 

Halperin, Rhoda 

1977 "Conclusion: A Substantive Approach to Peasant Livelihood." In Peasant Liveli- 
hood, pp. 269-297. Edited by Rhoda Halperin and James Dow. New York: St. 
Martin's Press. 

Hau'ofa, Epeli 

1978 "The Pangs of Transition: Kinship and Economy in Tonga." Australian and New 
Zealand Journal of Sociology 14:160-165. 

Holmes, Lowell 

1971 "Samoa: Custom Versus Productivity." In Land Tenure in the Pacific, pp. 91- 
105. Edited by Ron Crocombe. Melbourne: Oxford University Press. 

Kaeppler, Adrienne L. 

1971 "Rank in Tonga." Ethnology 10: 174-193. 

Kennedy, T. F. 

1958 "Village Settlement in Tonga." New Zealand Geographer 14: 161-172. 

72 Pacific Studies, Vol. 10, No. 2— March 1987 

Latukefu, Sione 

1967 "The Influence of the Wesleyan Methodist Missionaries in the Political Develop- 
ment of Tonga, 1826-1875." Ph.D. dissertation, Australian National University. 

Lingenfelter, Sherwood G. 

1977 "Socioeconomic Change in Oceania." Oceania 48: 102-120. 

Lockwood, Brian A. 

1971 Samoan Village Economy. Melbourne: Oxford University Press. 

Marcus, George 

1978 "Land Tenure and Elite Formation in the Neotraditional Monarchies of Tonga 
and Buganda." American Ethnologist 5:509-534. 

Martin, J. (for Mariner) 

1817 An Account of the Natives of the Tonga Islands, in the South Pacific Ocean. 3d 
Edition. London: Hurst, Chance and Co. 

Maude, Alaric 

1965 "Population, Land and Livelihood in Tonga." Ph.D. dissertation, Australian 

National University. 
1973 "Land Shortage and Population Pressure in Tonga." In The Pacific in Transition, 

pp. 163-185. Edited by Harold Brookfield. New York: St. Martin's Press. 

Morton, Keith L. 

1972 "Kinship, Economics and Exchange in a Tongan Village." Ph.D. dissertation, 
Anthropology Department, University of Oregon. 

1978 "Mobilizing Money in a Communal Economy: A Tongan Example." Human 
Organization 37:50-56. 

Nayacakalou, Rusiate R. 

1959 "Land Tenure and Social Organization." Journal of the Polynesian Society 68:91- 

Oliver, Douglas L. 

1961 The Pacific Islands. Garden City, New York: Anchor Books, Doubleday. 

Rousseau, Jerome 

1978 "On Estates and Castes." Dialectical Anthropology 3:85-96. 

Rubel, A. J., and H. J. Kupferer 

1968 "Perspectives on the Atomistic-Type Society— Introduction." Human Organiza- 
tion 27 -.189-190. 

Sahlins, Marshall D. 

1958 Social Stratification in Polynesia. Seattle: University of Washington Press. 

1965 "On the Sociology of Primitive Exchange." In The Relevance of Models for Social 
Anthropology, pp. 139-236. ASA Monographs No. 1. London: Tavistock Publi- 

1972 Stone Age Economics. Chicago: Aldine. 

Weber, Max 

1950 General Economic History. Translated by F. H. Knight. New York: Free Press. 



Darling Downs Institute of Advanced Education 

Toowoomba, Queensland 

Seaborne trade in the Pacific last century depended on shipping, capi- 
tal, connections, and markets. So crucial were these impersonal forces 
that they encourage an economic determinism by diverting attention 
from the human dimension of the business. In particular it is easy to 
overlook those who assumed the roles, so to speak, of hewers of wood 
and drawers of water. Seaborne trade, in short, was more than a matter 
of "prices, percentages, competitors and navigation" (Firth 1978:130); 
it was equally a matter of deckhands afloat, copra cutters ashore, and 
resident traders scattered throughout the archipelagoes. To put it 
another way, the interarchipelago trade in copra, beche-de-mer, and 
pearlshell could never have been an economic proposition unless the 
trading vessels involved had been manned by "native" crews who were 
cheap to maintain and feed; or had not thousands of Islanders through- 
out the Pacific cut the copra that those vessels carried away to distant 
markets. Nor could the copra trade have been conducted in the absence 
of the underrated yet ubiquitous European resident trader, sometimes 
trading on his own account but more commonly in company employ- 
ment, for whom the Pacific became "home." 

This breed of men has been pushed to the farthest margins of Pacific 
historiography. They seldom penned their memoirs 1 and have never 
received preferential treatment from historians. Furthermore, their 
marginal status within the European trading system has obscured, then 

Pacific Studies, Vol. 10, No. 2— March 1987 





= '• 


» oo- 

TO «- 

I 1 ! 

co co 
o o 



C> c 

oj to 

TO j 1 



I \ 




1 > 







— ' b_ 





Z O* 





. TO 





















< . 

1 1 — 






_J • 

LU. ' 

<°« 1 

„ - 

o s " e>.« .. 

h z - •■ . 

•' . • o 




1 ° 

3 s- 

® ai/H ajeo / 





■ 5° X 




1 ,A 

/ — 

w w 




. • 






2 *■ 








2 < Z ' 
-Or' < 1 

■ "// l 




. .^ s! 5 




< O -1 

• r-> 5 " <— — ■ 





1 yvi £z ^/ 8 

Resident Traders in Tuvalu 75 

and now, their crucial role within it. While other aspects of trading 
have become better understood in recent years, our knowledge of the 
resident outstation traders remains much as it was two decades ago 
when Harry Maude and Alastair Couper first drew attention to them 
(Maude 1968:270-273; Couper 1967). The only significant advance has 
come from Francis Hezel, who gives resident traders prominence in his 
history of the precolonial Caroline and Marshall Islands (Hezel 1983). 
Hezel, however, views the typical trader's life more optimistically than 
I am able. Given these circumstances, a deeper study on this social 
group needs little justification; what follows is a move in that direction. 
Rather than being a series of individual biographies, this paper brings 
together the fragmentary evidence in an attempt to construct a social 
profile, or group biography, of those traders who took up residence in 
the tiny island world of Tuvalu. 2 

Wider Background 

The nine coral atolls and reef islands of the Tuvalu group have always 
been peripheral to European interests. Small in size and modest in 
resources, they offered no scope for large-scale European settlement or 
plantation economies. 3 Trading contacts between Tuvaluans and Euro- 
peans initially occurred in 1821 when Captain George Barrett of the 
Independence II, the first whaling ship to enter Tuvaluan waters, sent 
parties ashore at Nukulaelae to obtain provisions for sick members of his 
crew (Stackpole 1953:279-280). Over the next forty years the most com- 
mon sail on the horizon was that of the broad-beamed whaler, over- 
whelmingly from Massachusetts. Although Tuvalu was located on the 
western edge of the On-the-Line whaling ground, whaling captains 
tended to avoid this badly charted, low-lying archipelago where the 
winds and currents were tricky, where local resources offered little 
scope for repairs and reprovisioning, and where whales were not partic- 
ularly abundant. Tuvalu, in other words, was more a thoroughfare than 
a whalers' resort, with many captains passing through and avoiding all 
chances of contacts with shore. Instead, Tuvaluans took the initiative, 
coming out in their canoes to exchange coconuts and mats for metal 
fishhooks, knives, glass bottles, and beads. Despite the intermittent 
nature of these encounters, trading between Europeans and Tuvaluans, 
as far as the latter were concerned, had begun in earnest, thus paving 
the way for less transient types of commercial contacts. 

By the 1860s the whalers' day in Tuvalu was all but over. The deple- 
tion of sperm whales combined with developments in the United States 

76 Pacific Studies, Vol. 10, No. 2— March 1987 

(civil war, gold rushes, and the substitution of whale products) resulted 
in whalers largely abandoning tropical latitudes. Their replacement 
was already on hand; during the previous decade the itinerant whaler 
had gradually been replaced by another breed of merchant— the 
seaborne trader. 

The first "traders proper" to enter Tuvaluan waters were freelance 
skippers in search of speculative cargoes. An early example involved the 
California vessel Rodolph on a Pacific cruise to gather produce for the 
San Francisco market (Kemble 1966:140-147). Trading contacts gradu- 
ally became more durable with the entry of the Sydney firms of Robert 
Towns and Company, J. C. Malcolm and Company, and Macdonald, 
Smith and Company, who pioneered the coconut-oil trade in the group. 
Initially their involvement in coconut oil was of a speculative nature, 
with their vessels dealing directly with Islanders and gathering cargoes 
wherever they could on an ad hoc basis. But they operated on a far 
larger scale than the independent freelance skippers: Tuvalu formed 
only one small part of their overall operations. Towns, for example, 
regarded the group as no more than an offshoot of his activities in Kiri- 
bati which, in turn, were but a sideline to his involvement in sandal- 
wood trading in Vanuatu. Periodically one of Towns' vessels was 
diverted from Vanuatu to pick up the coconut oil collected by his agents 
stationed in these remote archipelagoes (Maude 1968:263-267; Shine- 
berg 1967:108-118). Malcolm too had been active in Vanuatu, but had 
terminated this interest in about 1860 and redirected the Pacific side of 
his affairs to coconut oil. His ships plied Rotuma, Kiribati, and Futuna 
as well as Tuvalu, and he established a head station at Rotuma and 
posted a resident agent at Maiana in northern Kiribati (Macdonald 
1982:27; Maude 1981:81). Thus, from the onset Tuvalu was of little 
importance in trading company calculations and assumed significance 
only as part of an extended network of trading stations involving other 
archipelagoes as well. 

In order for the Island trade to function on such a scale, the trading 
companies had to rationalize their procedures. This was accomplished 
in two stages. First, the itinerant trading contacts between ship and 
shore that characterized the whalers and earliest traders gave way to 
the establishment of the island base, such as Malcolm's at Rotuma, from 
which company ships could strike out for neighboring archipelagoes. 
From there it was a short step toward developing networks of outsta- 
tions, each manned by a resident company agent (Couper 1967:51-52). 

This became the organizational form for trading throughout the 
Pacific. It was a conscious response to economy and efficiency and was 

Resident Traders in Tuvalu 77 

geared to minimize the difficulties of conducting seaborne trade in an 
extremely dispersed geographic setting. Resident traders served to stim- 
ulate local production by being on hand to barter goods for oil, which 
they would store until it was collected by the company ship. They also 
made possible the optimum utilization of ships' time, which was crucial 
to the fortunes of trading companies. This was more efficient than 
itinerant trading for speculative cargoes, where the vessel would lie idly 
at each point of loading while the crew went ashore for a cargo, quite 
possibly offending local sensibilities in the process. Not only was a ship's 
time, in a sense, irreplaceable, but ships themselves were costly items of 
capital equipment that depreciated rapidly and operated at high cost. 
Hence the need for their continual utilization and, in consequence, the 
rationale for resident traders. 

The Earliest Traders 

The first resident traders in Tuvalu arrived sometime during the 1850s 
as agents of Robert Towns (Maude 1968:265n). They were followed by 
others, such as Jack O'Brien at Funafuti, Charlie Douglas at Niutao, 
Peter Laban at Nukulaelae, and, perhaps, the two unnamed white men 
who went out to greet the New Bedford whaler Elizabeth at Nui in 
1861. 4 It is sometimes uncertain whether these individuals actually 
arrived as traders or whether they were deserters from whaling ships 
who sooner or later engaged as shore-based company agents or as inde- 
pendents gathering consignments of coconut oil for passing trading ves- 
sels. Jack O'Brien, reputedly the son of a New South Wales convict, and 
Charlie Douglas came to the Pacific as whalers, made the transition to 
beachcombing, and soon after turned to trade (Dana 1935:246-248; 
Restieaux MSa) . Then there is the case of John Daly, one of Towns' trad- 
ing captains, who was separated from his vessel by extraordinary misad- 
venture at Niutao in 1868: when towing twelve empty hogsheads ashore 
he drifted over the horizon, unnoticed by his crew, and drifted to Nanu- 
mea. He was rescued eight months later by a passing vessel. 5 But what- 
ever their reason for being in Tuvalu, resident Europeans who engaged 
in trade were advantaged by their linguistic monopoly: only a handful 
of Tuvaluans had ever enlisted on whaling vessels and few, if any, could 
speak English with sufficient fluency to act as middlemen in the passing 
trade in coconut oil (Munro 1985). 

The remarkable feature of these earliest traders was not their com- 
mercial significance but their religious impact in that they paved the 
way for the missionaries who followed (Brady 1975: 143n). Their 

78 Pacific Studies, Vol. 10, No. 2— March 1987 

motives were varied. There were the self-styled missionary traders, men 
such as Tom Rose at Nukulaelae and Robert Waters at Nui, who actively 
proselytized. Rose held rudimentary services on Sundays in response to 
the peoples' desire to know more about Christianity (Mrs. Chalmers 
1872:147-148). So did Waters, but with an eye to economic advantage: 
capitalizing on the temporal power that accompanied his assumed mis- 
sionary role, he instituted a system of fines payable only in coconut oil 
(Murray 1876:391-392, 409-410). Other traders were openly contemp- 
tuous of the pagan system. Charlie Douglas at Niutao and another 
trader at Vaitupu, known in oral tradition as "Titi," 6 set fire to all the 
religious structures (Alefaio 1979; Dana 1935:247-248). In dramatic 
fashion Jack O'Brien did the same at Funafuti, "not from any religious 
purpose, but because the ancient religion took up much of the time 
which he thought, rightly or wrongly, should be given to collecting 
copra [should be coconut oil] for him" (Sollas 1897:354-355). The fact 
that such actions failed to attract divine retribution was probably one 
factor in the eventual eclipse of the pagan religion. Even those traders 
who had no intention of eroding the foundations of the pagan religion 
usually had the effect of doing so (in the southern islands at least) for 
they came not as an isolated force but as part of an overall alien culture 
that the Tuvaluans perceived as being more powerful than their own. 
The Tuvaluans observed the worldly wealth, the impressive technology, 
and the literacy of these sporadic earliest Europeans, whether ashore or 
afloat; they could scarcely afford to ignore a god who so endowed his 
adherents. 7 Even such reprehensible actions as those of Tom Rose and 
Jack O'Brien in assisting the Peruvian slavers at Nukulaelae and Funa- 
futi respectively in 1863 (Graeffe 1867:1162-63) had the unexpected 
short-term effect of strengthening the appeal of Christianity. The survi- 
vors turned to the powerful new religion to restructure their communi- 
ties from the ruins of the immediate past (Maude 1981:174-175). 

By the mid-1860s most, perhaps all, the Tuvalu islands had played 
host to a resident trader. That decade was one of transition in the group. 
The trade in coconut oil was being phased out in preference for the 
more profitable copra. The London Missionary Society (LMS) also 
began stationing Samoan pastors throughout Tuvalu, and by degrees the 
group became a Protestant stronghold. Once established, the LMS 
entered the local trading relationship. Annual donations were solicited, 
thus diverting some of each island's resources away from the trader. At 
the same time, however, the LMS stimulated trade by creating a vigor- 
ous local demand for clothing, stationery, and building materials for 
churches. In this way the established two-way trading pattern between 

Resident Traders in Tuvalu 79 

traders and Tuvaluans was transformed into an interrelated three cor- 
nered relationship involving the LMS as well (Munro 1982:207-209, 

An additional change resulted from the "emergence of a new eco- 
nomic milieu in the archipelagoes" of the Pacific (Couper 1967:73), 
dominated by large, diversified, heavily capitalized companies quite 
different in character and objectives than their smaller rivals, whom 
they eclipsed, and better able to make money out of the dispersed island 
trade. In the forefront was the Hamburg firm of J. C. Godeffroy und 
Sohn, which exercised a near monopoly over copra trading in Tuvalu 
during the 1870s. The company placed traders on most islands of the 
group, including such individuals as Harry Nitz at Vaitupu and Martin 
Kleis at Nui. These men remained the rest of their lives on their respec- 
tive islands, marrying into the community and raising large families. By 
the following decade, however, Godeffroys' successor, the Deutsche 
Handels-und Plantagen-Gesellschaft der Siidsee-Inseln zu Hamburg 
(DHPG), encountered severe competition from H. M. Ruge and Com- 
pany, another German firm, and from Henderson and Mcfarlane of 
Auckland. By this time the trader's house, store, and copra shed in 
Tuvalu were as common a sight as the church, the maneapa (public 
meetinghouse), and the adjacent village green. 

Despite their diversity of social backgrounds, the overwhelming 
impression is that traders were a group of men, dissatisfied and often 
unsuccessful in other walks of life, who found a refuge on the margins of 
the Island trade. They were social casualties by and large. Most went 
into trading in the first place as a last resort after drifting in and out of 
various occupations — and sometimes in and out of trouble as well — in 
various parts of the world, ranging from goldmining in New Zealand, 
bushranging in Australia, fighting in a South American revolution, or 
service in the Hong Kong police force or the Royal Navy. Typically 
unmarried, but sometimes turning their backs on an unhappy mar- 
riage, they also leave the retrospective impression of weariness with the 
uncertainties of a wandering existence. In this gradual process of an 
unsettled disposition giving way to the desire for a more stable existence 
they drifted, often unexpectedly, into the Island trade, sometimes first 
as seamen but finally as company agents ashore. There is the suggestion 
too that one attraction of the trade was the absence of constant and 
direct supervision. Although more settled in their occupation than 
before, traders still tended to be men on the move, wearying of one 
island (sometimes driven away by the inhabitants) and moving on to the 
next with their dreams of sudden wealth receding ever further. Thus, 

80 Pacific Studies, Vol. 10, No. 2— March 1987 

most of the copra traders who came to Tuvalu were already identities on 
other archipelagoes and known by name at least throughout the Island 


The number of traders in Tuvalu probably exceeded seventy. I have 
been able to document the presence of sixty- two individuals, but there 
would have been others, particularly during the period before 1865, 
who are not mentioned in the sources. 8 The number of traders fluc- 
tuated, with some islands receiving more than others. On the drier reef 
islands of Niutao and Nanumanga, whose commercial potential is mea- 
ger even by Tuvalu standards, there was seldom more than a solitary 
trader, and indeed Nanumanga went for long periods without a trader 
at all. The atolls, by contrast, experienced almost continuous occu- 
pancy, often two or more traders at any given time, but there is no set 
pattern. The number of traders within the group also varied over time. 
At any given moment there could be as few as two or as many as four- 
teen, but usually there were between five and ten. The actual number 
reflected the overall activities of the trading companies involved. Dur- 
ing the 1870s there was an annual average of eight traders; most were 
employed by Godeffroys, who dominated the commerce of the wider 
region. During the 1880s, with increasing competition from other 
firms, the figure rose to ten but dwindled during the following decade 
to an average of five as one company after another terminated its inter- 
est in Tuvalu. Other companies filled the void, but the number of trad- 
ers in the group continued to decline because companies now preferred 
to deal directly with Tuvaluans; and by 1909 the remaining two Euro- 
peans in the group had long since abandoned their vocation to become 
"relics ... of a vanished class" (Mahaffy 1909; see also Wallin 1910). 

Relations with Other People 

Traders and Tuvaluans 

Whatever their initial diversity in social background, outlook, and tem- 
perament, traders in Tuvalu were soon forced into a common mold by 
the inescapable circumstances of their vocation, the atoll environment, 
and the people with whom they shared an island. At first sight it might 
appear that Tuvaluans were the arbiters of the trading relationship in 
that it revolved around their demand for trade goods and was depen- 
dent on the amount of copra they were prepared to produce in 
exchange. In point of fact the overall European trading system was not 
one in which Tuvaluans, or other Pacific Islanders, could participate 

Resident Traders in Tuvalu 81 

except at the bottom of a hierarchy of dependence. This limitation 
stemmed in part from cultural restraints imposed by kinship obliga- 
tions, community solidarity, and ethics of reciprocity that run counter 
to profit making and economic individualism. These same social obliga- 
tions mitigated against the success of indigenous commercial ventures 
based on the spirit of private enterprise. "In effect," as Couper explains, 
"a trading system based on cash tended to become inextricably en- 
twined with another based on kinship" (1967:125). Tuvaluans, more- 
over, lacked the necessary capital and connections and faced the hostil- 
ity of entrenched European interests whenever they chose to break into 
the new system of maritime commerce on its own terms. Thus, Tuva- 
luans and other Pacific Islanders were admitted only in a restricted 
fashion, providing they submitted to the trade's unilateral interests and 
pressures. They were the toilers — the copra cutters and deckhands — 
whose rewards were largely subject to external authorities and controls 
(see Couper 1968; Couper 1973). 

Resident traders were the point of contact between trading compa- 
nies and Tuvaluans, and the latter could exert telling pressure on the 
traders to get a better deal for themselves. The inherent conflict of 
interest between the two parties led to constant haggling over the price 
of goods and the value of copra. Boycotting the resident traders was a 
common enough occurrence on any given island as a means to raise the 
buying price of copra. Such embargoes on trade were invariably of 
short duration since the Tuvaluans in the meanwhile deprived them- 
selves of their only access to clothing, stationery, and other needed items 
from the trader's store, as well as cash for their missionary contribu- 
tions. Traders knew this and were wont to hit back by imposing trade 
embargoes of their own. Thus the missionary Nisbet discovered during 
the LMS's annual visit of inspection in 1875 that "The people [of 
Vaitupu] had fallen far short of their usual liberality in regard to the 
teacher's salary. He accounted for this by the fact that the stores had 
been closed, as the traders refused to comply with the demand of the 
people for an increase in the price of dried nuts" (Nisbet 1875:7-8). 

Another such incident occurred at Nanumanga five years later when 
Louis Becke closed his store following a dispute with a high-ranking 
Nanumangan. Being the only trader on the island Becke was in a strong 
position to set his own terms and conditions. It was not a situation 
where Islanders could play off rival traders against one another: all the> 
could do in the circumstances was to wait for trading vessels to call and 
conduct business with them. But no vessels came and finally the Nanu- 
mangans were reduced to pleading with Becke to resume normal trad- 

82 Pacific Studies, Vol. 10, No. 2— March 1987 

ing operations— or such at least is Becke's version of events (Becke 
1880). Disagreements over the quality or suitability of the nuts could 
also force trade to a standstill. When the traders at Niutao refused to 
purchase green nuts, which are useless for copra, the kaupule (council 
of elders) was "much aggrieved" and banned trade altogether on the 
grounds that the traders were acting "arbitrarily and unfairly" (Max- 
well 1881:5-6). 

Two separate issues were involved in the imposition of trade embar- 
goes by Tuvaluans. They were used both as a device to strengthen their 
commercial bargaining power and as a disciplinary measure against 
infractions of the local code. Captain Maxwell, who patched up several 
such disagreements during the cruise of HMS Emerald through Tuvalu 
in 1881, remarked that "the taboo is their only defence against any dis- 
honest trader, and their only means of enforcing good behaviour upon 
people towards whom they are not permitted to use force. Still there is 
no doubt that it may be, and sometimes is arbitrarily and vexatiously 
applied" (Maxwell 1881:5). Jack O'Brien's confrontations with the 
kaupule at Niutao, where he had gone to trade during the 1880s, pro- 
vides a case in point. He may well have been the sort of Irishman who 
felt "a sort of divine commission to fight against Kings and other rulers," 
but the kaupule were also dispensing justice rather too partially and 
"making laws having special reference to the poor Irishman whose iras- 
cible temper seems to have annoyed them" (Marriott 1883:17; Newell 

In other cases too the rights and wrongs of the matter were ill 
defined. When Thompson of Nui was fined for squabbling with his 
wife, he paid the fine but refused to appear before the chiefs for a man- 
datory scolding and instead remained inside his house behind locked 
doors. After the third summons the door was broken down, Thompson 
dragged bodily to the council house, and his residence robbed during his 
absence. Maxwell arbitrated and was told by another trader on the 
island that "the same thing would have been done to any native on the 
island; that the Kaipuli [sic] always enforced obedience to their de- 
mands according to the law; that Mr. Thompson was made aware of 
their laws when he first came . . ." (Maxwell 1881:3). In the end 
Thompson was quite amenable to reason: he wanted nothing more said 
about the stolen sixteen dollars, only an apology for being manhandled. 
Maxwell then suggested repayment of the stolen money and that 
Thompson's door be repaired, and finally Thompson and the "King" 
shook hands expressing their hope to "be better friends in future" (Max- 
well 1881:4). 

Resident Traders in Tuvalu 83 

The relationship between traders and local leaders embraced a range 
of situations and was fraught with tension. Even when disagreement 
with a trader was at an interpersonal or an interfamily level, it would 
ultimately have had to be resolved by the local leadership, since a trad- 
er's activities involved the island at large. Because each party was 
dependent on the other, it was mutually inconvenient for trading opera- 
tions to be suspended indefinitely, and disagreements were nearly 
always resolved sooner or later. In late 1885, for example, the council of 
Nanumea told George Winchcombe and Frank Jackson to leave the 
atoll the next time their firms' vessels called on the grounds that the two 
traders were "not good for the people." Jackson departed at the first 
opportunity, but Winchcombe stayed on and nothing more was said 
because, on sober reflection, the Nanumeans realized that they could ill 
afford to be without a trader on the island (Winchcombe 1881-1887:2). 

Traders and Missionaries 

Several other factors intervened to varying degrees to complicate trad- 
ers' daily lives. The Samoan pastors of the LMS frequently deployed 
their considerable local influence to damage the interests of a trader 
they happened to dislike, to the extent sometimes of instigating embar- 
goes against that trader. The visiting European missionaries received 
numerous complaints from traders that the pastors were interfering in 
trade or acting despotically, but almost invariably these charges were 
dismissed as being "trivial," "unfounded," or springing from jealousy of 
the pastors' local influence (Powell 1871:20; Davies 1880:6-7; Phillips 
1884:17; Winchcombe 1881-1887:31). Not prepared to be dictated to 
and loath to divert their busy schedules for the sake of aggrieved trad- 
ers, the visiting missionaries were also concerned to uphold the author- 
ity and standing of their outstation pastors. The European missionaries 
also tended to regard traders as "trashy whites" of godless deportment 
(although occasional friendships or regard for individual traders were 
also sometimes the case), and this likewise served to prejudice the recep- 
tion of even genuine complaints. Accordingly, the visiting LMS mis- 
sionaries perceived themselves not as the protectors of traders but quite 
the reverse: "The 'John Williams' has been a grand check on the doings 
of unscrupulous Traders," wrote one (Wilson 1886:14), while another 
reported that "as in other voyages so also in this one the evil example 
and influence of the traders scattered throughout these islands gives the 
Deputation no small amount of trouble" (Newel 1885:46). Traders' 
experience of Samoan pastor and European missionary alike largely 

84 Pacific Studies, Vol. 10, No. 2— March 1987 

explains why they were anti-missionary rather than irreligious as such, 
and Louis Becke had a point when he said that some traders were 
indeed "very religious men, although they don't show it" (Becke 

Some traders, however, did show their religious feeling but in their 
own way. Alfred Restieaux and George Westbrook on Funafuti so thor- 
oughly detested the island's dictatorial pastor that they refused to attend 
his church services and held their own private devotions on Sundays 
(Phillips 1881:8). Restieaux even refused to have the pastor baptize his 
children; he waited instead for the arrival of a warship and asked its 
chaplain to perform the ceremony (Maxwell 1881:2; Bridge 1883:2). 
Some other traders were more forthright in their efforts to undermine 
the pastor's standing. In 1890 Edmund Duffy arrived at Nanumea from 
the Fakaofo in Tokelau, where he had been at the center of a schism in 
the local church (Claxton 1889:9). At Nanumea he showed himself to be 
public spirited and directed the building of the road, which to this day 
runs on the ocean side of the main village (A. Chambers 1984:87-88). 
He also sided with the high chief against the pastor in a dispute over 
each other's sphere of authority, and when the pastor threatened and 
belittled the chief, Duffy translated a letter of complaint from the chief 
to the High Commissioner for the Western Pacific. 9 

Rarely did friendships develop between resident traders and the 
Samoan pastors of the LMS. Louis Becke, who traded at Nanumanga in 
the early 1880s, was later to idolize pastor Ioane, but on the grounds 
that he was different from the common run of pastors. Pearson 
(1970:240-241) has termed this attitude a "qualified racism" or the 
"cult of the exceptional Polynesian." Becke criticized the harsh and 
petty theocracies over which Samoan pastors typically presided, and 
then went on to say: 

But on this particular little island we had for our resident mis- 
sionary a young stalwart Samoan, who did not forbid his flock 
to dance and sing, nor prohibit the young girls from wearing 
flowers in their dark locks. And he himself was a mighty fisher- 
man and a great diver and swimmer, and smoked his pipe and 
laughed and sang with the people out of the fulness of his heart 
when they were merry, and prayed for and consoled them in 
their sorrow. So we all loved Ioane, the teacher, and Eline, his 
pretty young wife, and his two jolly little muddy brown 
infants; for there was no other native missionary like him in the 
wide Pacific. (Becke 1897:148-149) 

Resident Traders in Tuvalu 85 

Otherwise Becke so detested men of the cloth that he left Tuvalu in 1881 
for the Carolines, which were "free from that curse of the islands the 
missionary element" (Becke 1880). 

Most traders managed to maintain a reasonably polite relationship 
with visiting European missionaries, if only because they dealt with 
them so infrequently. The personalities of individual traders could also 
have a bearing on the outcome. Harry Nitz, the long-serving Godeffroy/ 
DHPG trader on Vaitupu, helped in the construction of the island's new 
church, which he then regularly attended in a manner befitting "a well 
conducted man" (Powell 1871:18; Davies 1873:4). Others, however, 
such as George Winchcombe, only served to confirm the missionary 
stereotype of the dissolute, worthless trader. As Louis Becke wrote 
of him: 

four years on Niutao and cannot yet talk the language in fact 
had to interpret for him. such a man to talk, my ears are actu- 
ally tingling now, I don't know how much more I would have 
suffered if it had not been for a case of gin I produced and by 
liquoring him up freely I got a little respite, he is a fair sample 
of too many island traders fond of liquor and never happy with- 
out some grievance against the natives, these are the men that 
give the missionaries such a pull over all traders — (Becke 1880) 

The irony is that Winchcombe "professed to be a cut above the ordinary 
trader." His airs and graces, his ostentatious use of long words, and an 
extreme possessiveness toward his wife resulted in Winchcombe being 
the butt of many unkind remarks (Dana 1935:254-256). At the same 
time he proved to be a sore trial to all he encountered. Fellow traders 
could not abide his selfishness: he was never prepared to reciprocate 
past kindnesses (Restieaux MSc). Nor did the chiefs on the various 
islands where he traded appreciate his litigious nature: he constantly 
appealed for their arbitration in his quarrels and then dissented from 
decisions that did not go his way. It is little wonder when he left Nuku- 
fetau that the people had no wish to accept another trader in their midst 
(Rooke 1886:10). 

Traders versus Traders 

As well as being at odds with representatives of the LMS, traders also 
had problems with each other and the companies to whom they were 
tied. Traders on the same island often provided each other with com- 

86 Pacific Studies, Vol. 10, No. 2— March 1987 

panionship, even if in different circumstances those men may not have 
been associates. But they were also certain to provide one another with 
competition and this could strain a friendship or even prevent one from 
developing, especially in the 1880s when falling world copra prices and 
increasing competition left many traders heavily in debt to their compa- 
nies. This competitive situation could result in price wars breaking out, 
and relations between traders on the same island then hit a very sour 
note. When Louis Becke arrived for a brief stay at Nukufetau in 1881, 
the only other trader on the atoll was George Winchcombe, whom he 
already disliked. Becke promptly raised the buying price of copra and 
lowered the selling price of his trade goods to the native producers, thus 
bringing Winchcombe's business to a halt (Maxwell 1881:3). 

The trading companies, moreover, did not always act in strict fairness 
toward their outstation traders, whom they considered to be insuffi- 
ciently honest, industrious, or sober. Feeling the pinch of hard times, 
Henderson and Macfarlane started charging their traders for shrinkage 
and debiting their accounts if, in response to competition, they raised 
the purchasing price of copra above a stipulated amount. (Restieaux 
MSb:12). In 1882 and 1883 this same firm was experiencing difficulty in 
keeping its far-flung and probably overextended trading network ser- 
viced, with the result that many of its traders, including those in 
Tuvalu, became 

completely destitute for stores, and even the necessaries of life, 
the vessels that should have supplied them being many months 
overdue. . . . One result of their being left in this distress is 
sometimes, that they are obliged to part with the produce they 
have collected for their own firm, in order to procure the neces- 
sary supplies, thereby gaining a character for fraudulent prac- 
tices which is not always deserved. (Le Hunte 1883:10-11) 

Yet "bondage" to an established trading company was, on the whole, a 
more secure arrangement than the earlier ad hoc procedure of collect- 
ing produce for free-lance trading-captains who might go out of busi- 
ness the next day and whose treatment often left much to be desired. A 
case in point is the treatment meted out to Winchcombe by the notori- 
ous Bully Hayes. Hayes landed Winchcombe at Nukufetau in 1872 but 
without provisions or suitable trade. Returning four months later Hayes 
invited Winchcombe and his Tuvaluan wife on board his ship, where he 
put a bottle of gin at the trader's disposal. The day ended with the hope- 
lessly drunk Winchcombe being tarred by the crew while Hayes was in 
his cabin with Winchcombe's wife where he "downed her on the sofa & 

Resident Traders in Tuvalu 87 

so forth"; finally, the two were dumped ashore and abandoned (Res- 
tieaux MSc:4-5, 7-10). 

Traders and Naval Captains 

In 1872, HMS Basilisk (Captain John Moresby) passed through Tuvalu 
and called at most islands. At Niutao Moresby warned that warship 
action would result "should they ever be unfriendly to white people." 
Two days later at Nanumea he heard that the local trader had been 
threatened, so a couple of shells were fired into the bush as a warning 
(K. Chambers 1984:110-111; Moresby 1876:79-80). Moresby's action 
was exceptional; never again did a British warship in Tuvalu waters fire 
a shot in anger. Traders eventually came to realize that they could not 
depend on naval protection. The first to do so was W. B. Thompson, 
who was fined and boycotted, and even occasionally assaulted, by the 
people of Funafuti during the mid- 1870s for arrogantly persisting with 
the notion that his Funafutian wife's family lands should be made over 
to him. Thompson regarded the matter as a test case and called for 
"powerful and vigorous action" to demonstrate once and for all that 
Islanders could not "with impunity insult, rob and committ [sic] Brutal 
Outrage upon a British subject." 10 

Although he was largely the maker of his own misfortunes, Thomp- 
son did have some valid complaints, though these were pointedly 
ignored by officers of both the Royal Navy Australian Station and the 
Western Pacific High Commission. Thompson's position as trader on 
Funafuti thus became quite untenable, and he had no option but to 
leave the atoll. Finally he took passage in 1878 on Henderson and Mac- 
farlane's schooner Belle Brandon commanded by Captain Frederick 
Ohlsen. It so happened that Ohlsen was embroiled in a separate trading 
dispute at Vaitupu where he had been menaced by a group of armed 
Islanders acting under instructions from the local Samoan pastor. When 
they eventually reached Fiji, Ohlsen and Thompson took their com- 
plaints to the acting High Commissioner for the Western Pacific, who 
dismissed them out of hand in abusive letters of reply. 11 

His verdict was predictable. The acting high commissioner was 
expressing what any other British colonial official would have said in 
the circumstances, if more vehemently. Similar sentiments were penned 
by a touring judicial magistrate who had 

not the least doubt in my mind that ... if Englishmen choose 
to settle in these spots — take native wives, — and identify them- 
selves with the natives, they must be satisfied to accommodate 

88 Pacific Studies, Vol. 10, No. 2— March 1987 

themselves to the rules and customs of the country, and not call 
for the interference of the captain of a man-of-war because they 
may be inconvenienced either in business or private matters by 
laws of general application to the whole people. . . . nothing 
to my mind can be worse than . . . sending Queen's ships to 
take up traders' squabbles, impose fines on whole communities, 
and, if necessary, enforce their payment by actual force, 
because — (in nineteen cases out of twenty) — some seedy loafer 
has, through his own acts, incurred the hatred of the islanders 
on whom he has been thrust, and who in enmity annoy him in 
order to drive him away. (Le Hunte n.d. : 14) 

In short, British official attitudes were little different from those of the 
LMS missionaries in that they disliked traders in general and resented 
being dictated to by them. Also at issue was the humanitarian principle 
that Islanders' rights were to be protected and upheld, and the practical 
consideration that imperial resources in the Western Pacific were 
patently insufficient to deal with more pressing matters, such as the reg- 
ulation of the labor trade in Melanesia, without a warship being 
diverted every time a "trader's squabble" was reported. Another colo- 
nial official summed up the situation exactly when he said, "if traders 
went in with their eyes open to these places merely for their own gain, 
they did not deserve the protection of their Government, and that it was 
better for the Governemnt to say that they would not protect them 
rather than that they could not, which is in reality the case" (Romilly 
1893:151). Indeed naval captains' hands were largely tied since they 
had no jurisdiction over Islanders or non-British nationals alike. In the 
absence of serious and unprovoked violence against British traders, the 
naval officers could not act against the Tuvaluans, and in any case none 
of them (apart from Captain Moresby in 1872) showed any inclination 
to do so. Nor did they regard themselves being obliged in any way to 
give active support to British traders in the group. Certainly Captain 
Maxwell was willing enough to listen to any grievances that British 
traders and island leaders may have had against each other, but he 
treated the two parties differently: the chiefs were generally given 
advice that was in no way binding while the trader was "cautioned 
. . . that he was a British subject and was amenable to British law for 
his acts, whatever they might be, whether he were the instrument of a 
German firm or any other" (Maxwell 1881:4-5; see also Macdonald 
After the declaration of a British Protectorate in 1892, traders' activi- 

Resident Traders in Tuvalu 89 

ties came under the scrutiny, albeit irregular, of a resident commis- 
sioner. One early commissioner, William Telfer Campbell, despised 
traders and proceeded against them on the slightest provocation. He 
was prompted to intervene in 1896 when Richard Collins of Nukulaelae 
threatened to shoot Lapana, the chief, during a heated disagreement. 
Campbell was singularly unimpressed with Collins' explanations that 
Lapana was normally "my best friend," that he had only threatened to 
"shove" Lapana outside, not to shoot him, and that he had no ammuni- 
tion for his revolver. Intent on ridding the Protectorate of Collins, 
Campbell had him conveyed to Funafuti and demanded two sureties of 
fifty pounds sterling each, in default of which Collins would be 
deported to Fiji. The high commissioner, while acknowledging that 
Collins was a "meddlesome" fellow, decided that Campbell's punish- 
ment was excessive because Lapana had offered provocation and also 
because Collins had really "been deported for matters outside the 
record." The deportation order was rescinded. Collins then "posed as a 
reformed character" before the visiting missionary from Samoa and got 
a free passage back to Nukulaelae with his family in the mission ship 
John Williams (Mrs. David 1899:284). Nevertheless Campbell got a 
measure of revenge by issuing a liquor prohibition order against Collins 
which helped to quieten him down. 12 

British traders were vastly dissatisfied with their treatment by offi- 
cialdom and Louis Becke spoke for many when he remarked that "a 
man in the South Seas now might as well be a Chinaman as an 
Englishman — for all the protection he will receive" (Becke 1880). It is 
not surprising in the circumstances that traders of British nationality 
sometimes toyed with the idea of changing their citizenship. 

The German navy, by contrast, was geared to the protection of Ger- 
man commerce in the Western Pacific and spent an estimated fifteen 
million marks doing so between 1875 and 1895 alone (Kennedy 1974: 
106). As W. B. Thompson said, by way of reproach, "The German 
traders are always . . . taunting us about the action of our Ships of War 
as compared with theirs and it is a fact that natives will insult an 
Englishman when they will not dare insult a German or American." 13 
Even so, German traders in Tuvalu were, in practice, no better placed 
than their British counterparts. Certainly the Ariadne called at Funa- 
futi and Vaitupu in 1878 and Captain Werner imposed trade and friend- 
ship treaties giving Germany most-favored-nation treatment. He also 
warned the chiefs at both islands that disruptions to German trade and 
shipping would no longer be tolerated, and at Vaitupu he intervened on 
behalf of the local DHPG trader, Harry Nitz, in a dispute over a piece of 

90 Pacific Studies, Vol. 10, No. 2— March 1987 

land (Werner 1889:320-330). But, as Captain Maxwell predicted, the 
Germans would never succeed in imposing their will on Tuvaluans in 
the absence of "strong and incessantly applied pressure" (Maxwell 
1881:5). The necessary follow-up action was not forthcoming and 
the only subsequent German naval visit occurred in 1883 when the 
Hydne called at Funafuti. In short, the Tuvalu archipelago was too 
unimportant both from a trading and labor-recruiting point of view 
to justify the regular oversight of the German navy, and without 
this coercive presence the treaties of trade and friendship fell away to 

Maxwell, in 1881, was the first naval officer, either British or Ger- 
man, to visit most Tuvalu islands since Moresby's 1872 cruise in the Basi- 
lisk. Ample time had passed for trading disputes to accumulate and 
intensify; indeed, Maxwell was dispatched to the Tuvalu, Kiribati, and 
Marshall islands precisely because there were so many outstanding mat- 
ters requiring official investigation. 14 Fair-minded though he was, Max- 
well left no doubt that the Royal Naval vessels were scarcely at the beck 
and call of traders. Successive visiting naval officers adopted the same 
attitude. The balance of local power now tipped in favor of the Tuva- 
luans and the quality of race relations improved as traders made greater 
efforts to keep on good terms with Tuvaluans. More aware than before 
that their business depended on local goodwill, most traders soon ac- 
cepted, if somewhat pragmatically, that the ground rules had changed 
and behaved accordingly. 

Naval activity in Tuvalu, then, helped the individual island com- 
munities in maintaining a show of integrity in the face of pressures from 
traders on the spot. But the naval captains were only one element in 
these local dramas; they combined with Samoan pastors, European mis- 
sionaries, and the competitive trading situation generally to strengthen 
the position of Tuvaluans in their dealings with their local company 
traders and, by extension, the wider European trading system itself. 

The Routine of Daily Life 

As well as having seemingly every other man's hand turned against 
them, the resident traders also had to contend with the realities of an 
unamenable physical and social environment. There were, admittedly, 
some compensations. At least the mosquitoes were not malarial. Fur- 
thermore, Tuvalu was an uncommonly peaceful place, and the only 
traders to be physically harrassed were those who brought it upon 
themselves, like W. B. Thompson at Funafuti. But the limitations of 
the restricted atoll diet and the doubtful nutrition of the provisions from 

Resident Traders in Tuvalu 91 

company ships, if they came at all, made inroads into the traders' 
health, lowered their resistance to secondary illnesses, and sapped their 
vitality. When Robert Louis Stevenson and his entourage visited several 
islands of the group in 1890, every trader they met was in poor health, 
whether it be from food-related complaints such as anemia and boils or 
from other ailments such as elephantiasis. The two traders at Funafuti 
were described as "wretched looking objects," and Stevenson's wife, 
Fanny, was dismayed when the leprosy-inflicted trader at Niutao, 
whose "fingers were dropping off," shook hands with her (Mrs. Steven- 
son 1914:89-106). Moreover, Western medical facilities were nonexis- 
tent ashore and recurring illness was a fact of life among traders. On 
one occasion at least the timely arrival of the missionary barque John 
Williams probably saved the life of a sick trader (Powell 1879:3); and in 
1896 Harry Nitz of Vaitupu sought treatment in Fiji for his skin com- 
plaints, said to be leprosy. 15 

Socially and intellectually traders were little better off. Theirs was an 
isolating vocation. Detached from the mental climate that had shaped 
their outlook and values, they were now transplanted in a markedly dis- 
similar social framework where the dull routine of village life combined 
with the sameness of the scenery and the infrequency of diversions 
served to depress the senses, impose a tedium on their lives, and encour- 
age a pattern of heavy drinking. Nor was a trader's existence enlivened 
by the spartan simplicity of his dwelling, typically a wretchedly 
appointed native-style house largely bereft of creature comforts. As 
George Westbrook pointed out after several years at the game in Funa- 
futi during the 1880s: 

If you would only bear in mind what a wretched life it is living 
on one of these sandbanks, no company, no amusements, no 
Theatres, no Bank Holidays, no beefsteak or fresh vegetables 
for 7 years, if sick no doctor, no news from home or friends, let- 
ters often lost or laid carelessly by, several times I have not 
received letters until long after written. 16 

Outside observers would not have been surprised at such regrets. "It is 
one of the saddest features of this . . . voyage [to Tuvalu, Tokelau, and 
Kiribati]," reported one LMS missionary more in pity than disgust, "to 
note the solitary and wasted lives [of traders] on almost every island, 
awaiting a lonely grave in this vast ocean solitude" (Claxton 1889:9). 
"Truly," observed another traveler, "the traders life on these islands 
must be fearfully monotonous. Some are unvisited for nine months or a 
year and the natives are far from cheerful company for an educated 

92 Pacific Studies, Vol. 10, No. 2— March 1987 

man" (Woodford 1884:75). Actually most Tuvalu islands experienced a 
far higher frequency of shipping contacts by the 1880s as a direct result 
of increasing competition in the island trade. Passing ships sometimes 
left behind reading matter (Gill 1872:9; Le Hunte n.d.:15; Woodford 
1884:16; Thurston 1893:10), and the arrival of a company vessel was 
usually an eagerly awaited event as it brought provisions, mail, news of 
the world, and fellowship (Dana 1935:197-199). But the diversion was 
only momentary. Often before nightfall the ship had disappeared over 
the horizon, leaving the traders to resume their monotonous, unhealthy, 
and enervating existence. 

Part of the problem was the ambivalent social identity that Tuvaluans 
accorded traders in their midst. Whether trading on their own account, 
or in the capacity of company agents, traders were never just individu- 
als to be judged on their own merits. They were also seen as repre- 
sentatives of an alien trading system whose rationale violated a local 
reciprocity system based on generosity and sharing. The disjunction 
between the two was fundamental, and being associated with the for- 
mer resulted in traders being part of, yet detached from, village life. 
Marrying into the community, as most traders did, helped ease these 
problems of allegiance and identity; but marriage could be a two-edged 
sword. It gave traders access to a domestic life and the support, more or 
less, of an extensive network of af fines. But this wider social identity 
could also carry the penalty of the wife's relatives expecting preferential 
treatment in trading relations. Nevertheless, these marriages were usu- 
ally lasting relationships from which some of the leading families in 
Tuvalu— the O'Briens, the Kleises, and the Restures (a corruption of 
Restieaux) — are descended. 17 

The harsh facts in a trader's life were the heat, tedium, and mos- 
quitoes, the often claustrophobic social pressures of village life, and the 
inability to get away from one's problems. The pervasiveness of mission- 
inspired local laws and pastor domination were another bone of conten- 
tion; often small in themselves, such irritations had a cumulative effect 
and intruded heavily on a trader's existence. Sometimes the sum total of 
frustrations and hardships became intolerable and, provoked once too 
often, traders could react dramatically. The Chinese trader at Niutao in 
1878 reached the limit of his endurance when he was fined for killing a 
chicken on Sunday. In a fury "he killed it a second time! ! !" and when he 
was dragged off to the maneapa he used the "most filthy language 
towards the King and chiefs" (Turner 1878:56). 

Given the frustrations of their lives and the basic conflict of interests 
between traders and Tuvaluans it is not surprising that an undercurrent 
of racial antipathy was often part of a trader's stock of attitudes, even 

Resident Traders in Tuvalu 93 

though most were married to a Tuvaluan. Much of the time such senti- 
ments were well under control since their business operations in large 
part depended on local goodwill. But in the case of an open disagree- 
ment and particularly a boycott, these suppressed feelings of dislike and 
contempt were liable to come explicitly to the fore. 18 

The paradox is that station-trading became a way of life. Frank Tho- 
mas had been twenty-seven years in the business before he came to 
Vaitupu in 1882. Normally he was an independent but on that occasion 
he was in the employ of Henderson and Macfarlane (Bridge 1883:3; 
Dana 1935:201-207; Davis 1892:81-82). Another familiar figure in the 
Island trade was the "old man-of-war's man" Charlie Douglas, who for 
over thirty years following his departure from Niutao was to be found 
on one or other of the Marshall Islands until his death in 1892 from a fall 
on board a visiting vessel (Young 1875-1877:26; Le Hunte 1883:44; 
Davis 1892:21,92). Tom Day (or O'Day) was another permanent fix- 
ture. It was for his benefit that Captain Moresby bombarded Nanumea 
in 1872. He left the atoll soon after and at one stage was to be found at 
Nikunau in southern Kiribati. Described as "another of these pitiful old 
blackguards," he returned to Tuvalu in 1893 as Henderson and Mac- 
farlane's trader at Nanumanga and died the following year at Na- 
numea. 19 

There were a number of reasons why traders remained traders. Trad- 
ing life was usually no worse and often a great deal better than anything 
available elsewhere. Many, it will be recalled, were traders to begin 
with because they could not make a success of anything else. Those who 
attempted to break out of the system were usually forced back into it 
again. As time went on their options contracted and their life chances 
diminished. Often no richer than the day they started trading and in far 
worse physical shape, they had little prospect of gaining a livelihood 
back in Europe or Australia, however much they yearned to go back 
and settle down with their "own people" (Dana 1935:260-267). Ideas of 
return were doubly remote for traders who had married across racial 
lines and who knew that their wives and children would have difficulty 
adjusting and would never be socially acceptable outside the Pacific. A 
certain ambivalence may also be detected because the Island world and 
the European world each had their attractions and drawbacks. In the 
end it was a choice between either one or the other, and circumstances 
usually forced a return to the Islands, where at least the trader was his 
own master in the limited sense that his work was not subject to con- 
stant oversight. Jack Buckland, who traded at Niutao and Nanumea 
during the 1890s, made the most of his situation. In singular fashion he 
"spent a short period each year in Sydney playing spendthrift on the 

94 Pacific Studies, Vol. 10, No. 2— March 1987 

accumulations of a small funded income and the rest of the year vege- 
tating penniless as a petty trader out in the islands." 20 

Once they had accepted the reality of their situation, traders often 
then made a virtue out of necessity and developed an ethos and mys- 
tique of their own in which they became the sturdy and self-reliant 
knights of commerce. The reality was that traders were uncommonly 
dependent— on their company vessels for provisions, reading matter, 
and news of the outside world; on warships for all these things in lesser 
measure and for protection; and on Tuvaluans for copra. Station-trad- 
ing, moreover, was a dead end and so became a way of life by default. It 
was exceptional for a trader to find another livelihood. Louis Becke got 
out and eventually used his literary skills to make a living; even then he 
wrote mostly about the "world of traders, supercargoes and their native 
contacts" of which he was once part (Maude 1967:225). Otherwise, 
traders moved to port towns like Apia where their domestic arrange- 
ments were within the pale of civic respectability. 

Here we return to the wider question of resident traders' self-percep- 
tions and attitudes. On the subject of Louis Becke, Pearson (1984:81) 
makes the observation that Becke's writings show "a notable lack of crit- 
icism or even of the desire for any other kind of world than . . . one free 
from the restraints of home society, free from the operation of con- 
science." This statement indirectly points to the dilemma and the 
ambivalence of the traders' situation. They were disparaged by mis- 
sionaries, distrusted by their employers, discriminated against by naval 
authorities, hounded by resident commissioners, regarded as "seedy 
loafers" by a British judicial commissioner (Le Hunte n.d.:14), and con- 
demned by a high commissioner for introducing their "rotten, pestilen- 
tial civilisation" (Thurston 1893). In consequence, traders increasingly 
became estranged from their own society and developed something of a 
siege mentality. Left to fend for themselves, or so it seemed, they in turn 
saw no obligation to uphold the values of their own societies. Many 
therefore left the impression, often justified, of being profligate, god- 
less, hostile toward constituted authority, and devoid of enduring 
values. Traders thus acquired an equivocal social identity. On the one 
hand they were regarded by other Europeans as an outgroup. Yet they 
were more or less unsuccessful in integrating into village life and so they 
remained tied, if tenuously, to the moorings of the European world. 

The Decline of Trading Life 

In 1892 Britain declared a Protectorate over the then Ellice Islands. The 
move slightly postdated other changes that had gradually whittled 

Resident Traders in Tuvalu 95 

down the number of resident traders in the archipelago. This process of 
attrition began in 1888 with Ruge's bankruptcy and the final with- 
drawal of the DHPG a year and a half later, leaving Henderson and 
Macfarlane with the field to themselves (Munro 1982:198-205). Their 
monopoly was strengthened in 1893 by a mail subsidy for their vessel SS 
Archer to call at Fiji, Tuvalu, and Kiribati (Couper 1967:83). Capitaliz- 
ing on this concession, Henderson and Macfarlane landed a handful of 
new traders in the group, namely Jack Buckland at Niutao, Tom Day at 
Nanumanga, Richard Collins at Nukulaelae, and Edmund Duffy at 
Nanumea. 21 But Henderson and Macfarlane's monopoly did not last. 
The Australian company Burns Philp entered the Gilbert and Ellice 
Islands Protectorates in 1898 and consolidated its position in 1902 with 
the award of an Australian government mail subsidy to provide regular 
shipping to the Protectorates (Buckley and Klugman 1981:80-81). They 
in turn encountered competition from Levers Pacific Plantations in 
1903, 22 and from Captain E. F. H. Allen of the Samoa Shipping and 
Trading Company in 191 1. 23 

Several changes differentiated the new trading dispensation from the 
precolonial order. It was based largely on steamships rather than sailing 
vessels; Anglo-Australasian companies rather than German firms domi- 
nated the business; and the shore-based company agent gradually 
became redundant. Tuvaluans, now compelled to make quantities of 
copra for the Queen's Tax, had become so accustomed to the require- 
ments of trading companies that they dealt directly with the trading 
steamers' supercargoes. Indeed at Funafuti in 1910 a native trading 
company was running in direct competition with the resident half-caste 
trader, who was probably a son of Jack O'Brien (Buckley and Klugman 
1981:265; Macdonald 1982:141). A further structural change to trading 
in Tuvalu occurred in 1914 when Captain Allen transferred his Apia 
headquarters to Funafuti. 24 From this atoll base he plied Tuvalu and 
neighboring archipelagoes much in the manner of the speculative 
owner-traders of the mid-nineteenth century, and serviced plantations 
on the uninhabited islands of Nassau, Niulakita, and in the Phoenix 
group (Allen n.d.). In other words the copra trade in Tuvalu reverted 
back to a predominantly ship-based operation in which the shore-based 
agent of old had no place. 

Nevertheless, a handful of resident traders in the Protectorate re- 
mained on their islands, long-established identities who were so institu- 
tionalized that they stayed on to serve out their time if only in retire- 
ment. When Henry Nitz passed away at Vaitupu in 1906, and Martin 
Kleis at Nui in 1908, each had spent over thirty years on his respective 
island almost without break. 25 Jack O'Brien's death at Funafuti in 1899 

96 Pacific Studies, Vol. 10, No. 2— March 1987 

(Waite 1899:540n) ended a turbulent career in the group that spanned 
four decades. He arrived in pagan times and stayed on to witness many 
transformations in the Tuvaluan way of life. In his twilight years at 
Funafuti he "constantly expatiated on the good old times when he first 
came to the island, when the people held feasts, public games, dances, 
and such-like pleasures, most of which have been put down by the mis- 
sionaries. He said things were much more lively in those days" (Mrs. 
David 1899:167). Ironically, he himself had helped prepare the way for 
the LMS by desecrating and then destroying many of the old religious 
structures forty years before. 

The last of this group of stayers was Alfred Restieaux. He was typical 
of the trader who would die in the Islands, "perhaps cherishing to the 
last the fancy of a visit home" but "doomed ... to remain indefinitely 
on one narrow atoll" (Farrell 1928:351; Stevenson 1900:2). Like O'Bri- 
en he said that only his Tuvaluan wife and children prevented his return 
to somewhere like Sydney (Mrs. David 1899:132; Mrs. Stevenson 
1914:91). Both were probably rationalizing to a certain extent. Heavily 
in debt to the DHPG throughout the 1880s he was simply abandoned 
when they pulled out of Tuvalu. He never traded again but instead went 
to Nukufetau, his wife's home island, to live out his days in reduced cir- 
cumstances. His health was not good and his eyesight progressively 
deteriorated. Although described by visiting naval captains as "doing 
nothing," Restieaux in fact wrote a series of reminiscences that have 
proved valuable to historians and other students of this period in 
Tuvalu. With his passing in 1911, an era in Tuvaluan commercial his- 
tory drew to a close. 26 

Tuvalu was one of the first archipelagoes to experience the demise of 
its resident trader population. 27 By 1911 the rationale for trading com- 
panies placing agents on islands— namely to encourage production for 
barter and to reduce the turnaround time of the ship collecting the 
cargo — was no longer there. Production was now stimulated by a 
King's Tax, payable in copra, and the trading routine was so well estab- 
lished that Tuvaluans preferred to deal directly with the visiting super- 
cargoes. The middleman had been effectively cut out and in 1910 the 
two remaining traders in the group were the half-caste descendants of 
late traders (Wallin 1910) . 

Wider economic imperatives combined with local circumstance to 
preclude the return of resident traders to Tuvalu. Burns Philp estab- 
lished a headstation in Kiribati in 1913 that included Tuvalu and the 
Marshall Islands within its orbit. Although subsidized by regular 
renewals of its Australian government mail contract (Buckley and 

Resident Traders in Tuvalu 97 

Klugman 1983:25, 66, 99), the firm lost ground in Tuvalu and concen- 
trated instead on the more lucrative Marshall and Kiribati archipela- 
goes. The matter of establishing trading stations throughout Tuvalu was 
often discussed. But the cost of erecting such facilities would have out- 
weighed the benefits and it was recommended that the group could be 
better worked if another vessel was added to the Gilberts fleet. The 
tempo picked up during the early 1920s with company vessels making 
two or three trips to the group per year and a trading station being 
established at Nukufetau. However, as a result of diverting a vessel two 
or three times a year to service Tuvalu as well as the occasional call to 
Ocean Island, the seat of government of the Gilbert and Ellice Islands 
Colony, Burns Philp's Gilberts fleet became overworked. The firm 
attempted to rationalize by withdrawing from the Marshalls and trans- 
ferring SS Murua to the Tuvalu run, which resulted in the wreck of the 
vessel at Nanumea in April 1921. 28 

With his headquarters at Funafuti, Burns Philp's only competitor in 
Tuvalu, Captain Allen, was better placed to profit from the group. His 
advantage was increased by the success with which he tendered for the 
King's Tax, because he usually carried superior lines of trade goods 
(Couper 1967:103), and because his agreement with Burns Philp to sell 
trade goods at the same prices meant that he could not be undercut. 29 
He also ran profitable sidelines by providing building materials for 
churches and conveying government personnel. But during the 1920s, 
with a combination of family problems and his ships being wrecked or 
condemned, Captain Allen's enterprises fell on hard times and were 
bought up by Burns Philp shortly after his death in 1924. The following 
year the Tuvalu portion of Burns Philp's Tarawa Branch was transferred 
to the firm's Apia Branch. 30 

These organizational changes to the Tuvalu copra trade effectively 
put an end to any possibility of a recrudescence of resident traders in the 
group. Essentially, then, they were a transitional social group that 
existed over a fifty to sixty year period to meet the needs of a particular 
stage of the Island trade, dying out when that stage gave way to 
another. In their time they played an important part in the most 
momentous period of Tuvalu history, which saw the shift from an isolat- 
ed, independent, self-sufficient, and pagan life-style to one that was 
Christian, literate, partially dependent on a range of imported goods, 
involved in a wider sphere of political activity, and largely accepting of 
missionary and colonial rule. 

Yet the passing of this specialized social group was swift and almost 
unnoticed. In 1926, only fifteen years after Alfred Restieaux's death at 


Pacific Studies, Vol. 10, No. 2— March 1987 

Nukufetau, Tuvalu's commercial environment took another change in 
direction, and with the establishment of the first cooperative trading 
society (Macdonald 1982:142) the group's economic climate was even 
further removed from the one that had enabled the existence of resident 
traders. Such was the pace of change that by 1926 most of Tuvalu's 
younger generation would never have laid eyes on a resident trader. 


BP Becke Papers, Mitchell Library, Sydney. 

BPh Records of Burns Philp Company Ltd. , Rare Books and Spe- 

cial Collections Library, University of Sydney. (These rec- 
ords are presently being reboxed and to some extent re- 

CO 225 Records of the Colonial Office, Public Records Office, 

ML Mitchell Library, Sydney. 

PMB Pacific Manuscript Bureau, Manuscript Series. Available on 

microfilm at member libraries of the Bureau. For details see 
any recent issue of Pambu, the Bureau's newsletter. 

RNAS Records of the Commander-in-Chief, Royal Navy Australia 
Station, National Archives of New Zealand, Wellington. * 

RP Restieaux Papers, Alexander Turnbull Library, Wellington 

(consulted on microfilm in the Department of Pacific & SE 
Asian History, Australian National University, Canberra). 

SSJ Records of the London Missionary Society, South Sea Jour- 

nals. School of Oriental and African Studies, London. * 

SSL Records of the London Missionary Society, South Sea Let- 

ters. School of Oriental and African Studies, London. * 

SSR Records of the London Missionary Society, South Sea Rec- 

ords. School of Oriental and African Studies, London. * 

TP Towns Papers, Mitchell Library, Sydney. 

WP Westbrook Papers, Alexander Turnbull Library, Wellington. 

WPHC 4 Records of the Western Pacific High Commission, Series 4, 
Inwards Correspondence— General. Public Records Office, 

Consulted on microfilm at either the Mitchell Library, Sydney, or the National Library 
oj Australia, Canberra. 

Resident Traders in Tuvalu 99 


This paper is drawn from a Ph.D. thesis presented to Macquarie University, Sydney, and 
supplemented by subsequent research. I am grateful to my supervisor, Stewart Firth, for 
his advice and encouragement. 

Thanks are also due to Bruce Da we for casting his eye over an earlier version. A later 
version benefited from the detailed comments of Michael Goldsmith and Keith Chambers. 
Hugh Laracy, Ken Buckley, and Kristine Klugman gave privileged access to materials that 
form the basis of their unpublished research. Burns Philp allowed me to consult their 
records and the staff of the Rare Books and Special Collections Library at the University of 
Sydney, the custodians of these records, facilitated my research among them. The Institute 
Research Services of the Darling Downs Institute of Advanced Education financed the 
post-thesis research. 

Harry Maude has been unfailingly kind and supportive over the years. I dedicate this 
paper to him. 

1. Pacific Island traders were like English agricultural laborers and Australian convicts 
in that they left few written remains. Only four of the seventy or so traders in Tuvalu 
between c. 1855 and 1909 left reminiscences or letters. George Westbrook's experiences at 
Funafuti and Niutao during the 1880s have been published as Dana 1935:169-258. Fur- 
ther material is in the Westbrook Papers housed in the Alexander Turnbull Library, Wel- 
lington. The Westbrook Papers also contain a valuable set of manuscripts written by his 
fellow trader at Funafuti, Alfred Restieaux, available only on microfilm. They are not 
autobiographical but relate to the doings of other traders. The surviving letters and papers 
of Louis Becke are held by the Mitchell Library, Sydney. Becke also published a number of 
semi-fictional accounts, some of which have been utilized in the present article. Included 
among the Becke Papers is a useful diary kept by George Winchcombe, who traded on sev- 
eral of the Tuvalu islands. See the bibliography for details. 

Despite being a small sample, the situations and experiences of these four men are suffi- 
ciently representative of traders in Tuvalu that I feel safe in making extensive use of their 
writings. Inevitably, however, I have had to rely for most of my information on the writ- 
ings of people who were not traders. 

2. Tuvalu is the present-day name for the former Ellice Islands. The other indigenized 
place-names used in this article are Kiribati (which includes the former Gilbert Islands) 
and Vanuatu (formerly the New Hebrides). Unlike Kiribati and Vanuatu, Tuvalu appears 
to have been the traditional name its inhabitants applied to their island group. 

3. The single exception was the plantation at Nukulaelae operated by the Godeffroy/ 
DHPG establishment from 1865 until 1890 (see Munro and Besnier 1985). Unreferenced 
statements in this section relating to the European trading system and Tuvalu's place in it 
have been drawn from Brookfield 1972: chs. 1-4; Couper 1967; Macdonald 1982: ch. 2; 
Munro 1982: chs. 2, 7, 8. 

4. Restieaux MSa; Moresby 1872:163; Logbook of Elizabeth, 16 Sept. 1861, Old Dart- 
mouth Historical Society and Whaling Museum, New Bedford, Mass. (PMB 290, frame 

5. Eury to Towns & Co., 10 Feb. 1868, TP, Uncat. MSS set 307, item 89; Eury to Towns 
& Co., 19 Feb. 1868, TP, Uncat. MSS set 307, item 91; Sydney Mail, 28 Aug. 1869, 12b. I 
owe these references to Harry Maude. 

100 Pacific Studies, Vol. 10, No. 2— March 1987 

6. Titi's identity is uncertain. He may have been Robert Towns' agent Solomon Heather 
(Maude 1968:265m); or he may equally have been a man named Marshall who was dis- 
charged from the New England whaling bark Stafford in 1861 (see Logbook of Stafford, 2 
Nov. [1861], Old Dartmouth Historical Society and Whaling Museum, New Bedford, 
Mass. [PMB 957]). In 1874 a Mr. Marshall was trading at Vaitupu as Bully Hayes' agent 
(see Samoanische Zeitung, 11 Jan. 1914, lib). This was probably the same person who 
was discharged from the Stafford, but there is no evidence that Marshall lived contin- 
uously on Vaitupu between these dates. 

7. This interpretation is more fully argued in Munro 1982: ch. 4. On the subject of trader 
involvement in the destruction of the pagan religion, Michael Goldsmith (personal com- 
munication, 16 Aug. 1982) suggests that some of the "softening-up" they effected might 
have been because Tuvaluans did not regard missionaries and traders as belonging to dis- 
crete occupational categories. This is not to suggest that there were no differences between 
the two in the eyes of the Tuvaluans. Rather, the differences that the traders and missiona- 
ries considered important very likely did not match local criteria, with the result that trad- 
ers were perhaps seen as "missionaries as well," just as the missionaries-proper who fol- 
lowed would have been perceived to some extent as "traders." 

8. There is an annual island-by-island breakdown of traders in Tuvalu for the years 
1865-1892 in table 7:3 of my thesis (Munro 1982:186-190). I will be happy to respond to 
readers who require further details. 

9. Vaitoru to Thurston, 16 Nov. 1893, translated by E. A. Duffy, trader on Nanumea, 
WPHC 4, 76/1893. The background to this episode is detailed in Chambers 1984:112-113, 
151; Munro 1982:147, 147n. 

10. Thompson to Gorrie, [n.d.], end. in WPHC 4, 30/1878. 

11. See ends, in CO 225/1/16498 and in RNAS 13/49; Turner 1878:11-21. 

12. See ends, in WPHC 4, 69/1897; WPHC 4, 316/1897; WPHC 4, 456/1897; CO 225/52/ 
10715; CO 225/52/10179; CO 225/52/18023. 

13. Thompson to Gorrie, [n.d.], end. in WPHC 4, 30/1878. 

14. Gordon to Kimberley, 23 Apr. 1881, CO 225/7/9875; Maxwell to Gordon, 16 Apr. 
1881, WPHC 4, 80/1881. 

15. Campbell to Thurston, 5 Dec. 1896, WPHC 4, 68/1897. 

16. Westbrook to the trustees of Henderson & Macfarlane's estate, 10 Jan. 1890, WP, 
folder 43. 

17. At this point it is worth drawing attention to a comment by Niko Besnier (personal 
communication, 23 Jan. 1986). He suggests that 

the adaptational difficulties faced by traders on Tuvaluan atolls can probably be 
explained by a combination of the extreme isolation of the atolls and of another 
factor, namely the very big difference in ethos between Tuvaluan and Western 
cultures. Margaret Mead ("Public Opinion Mechanisms among Primitive Peo- 
ples," Public Opinion Quarterly 1 [1937]: 5-16) has suggested that cultures may 
be typologized in terms of the basis on which their ethos is channeled, recogniz- 
ing "personal" (e.g., Arapesh), "positional" (e.g., the Iatmul), "traditional" 
(e.g., Bali), and "positional-traditional" (e.g., Samoa, the Zuni) categories (cf. 
also Felix and Mary Keesing, Elite Communication in Samoa. Stanford: Stan- 

Resident Traders in Tuvalu 101 

ford University Press, 1956, p. 258, for discussion on Samoa). Tuvaluan culture, 
where many of the individual's choices in life are made by the social system, is 
positional-traditional with a vengeance. This accounts for (i) the receptiveness 
of Tuvaluans to the laws established by the Samoan pastors (and later Grimble) 
dictating behavior in excruciating detail of everyday life; and (ii) the extreme 
difficulties outsiders may have dealing with such ethos (even today, as I am sure 
you know well!). 

18. This aspect of life goes unmentioned in Becke's nostalgic published account of his 
eleven months' residence at Nanumanga (see Becke 1909:54-104). For Restieaux's opinion 
of Becke, see Restieaux MSd. 

19. Hayter 1871-1873: entry for 21 July 1872; Mrs. Stevenson 1914:120-123; Thurston 
1893:9-10; Swayne to Thurston, 18 Dec. 1893, WPHC 4, 21/1894; Swayne to Thurston, 
17 Jan. 1895, WPHC 4, 42/1895. 

20. Furnas 1951:365. There was an unhappy sequel, which Mrs. Stevenson (1914: 175n.) 
relates: "Some years ago when Jack was at his station he received word that his trustee, 
who was in charge of his property, had levanted it all. Whereupon poor Jack put a pistol to 
his head and blew out what brains he possessed. He was a beautiful creature, terribly 
annoying at times, but with something childlike and appealing — I think he was close to 
what the Scotch call a natural — that made one forgive pranks in him that which would be 
unforgivable in others. He was very proud of being the original 'Tommy Hadden' in 
[R. L. Stevenson's book] the 'Wrecker,' and carried the book wherever he went." 

21. Mrs. Stevenson 1914:106; Davis 1892:52; Swayne to Thurston, 18 Dec. 1893, WPHC 
4, 21/1894. Even so not all islands had a resident trader; in 1897 the resident commissioner 
reported, "There is no trader on the island [of Nukufetau] but I believe one of Messrs Hen- 
derson and Macfarlane's traders will shortly commence trading. His arrival is anxiously 
looked forward to by the great majority of the natives who are suffering from a tobacco 
famine." See Campbell to Berkeley, 22 Sept. 1897, end. in CO 225/52/25701. 

22. Samoanische Zeitung, 4 Nov. 1905, 8c, 2 Dec. 1905, 8b, 8 Feb. 1908, 8c; Wallin 1910. 
Ken Buckley and Kris Klugman kindly gave me access to Wallin's report of 30 Jan. 1910. 
Wallin wrote several other reports that Dr. Buckley and Ms. Klugman have edited for pub- 
lication as accompanying text for a photographic volume on Burns Philp's Pacific activi- 
ties. The book, to be published by George Allen & Unwin Australia, is scheduled for 
release in early 1987. 

23. Allen n.d. I am grateful to Hugh Laracy for making available a Xerox copy of this 
typescript. Dr. Laracy will be writing an essay on Captain Allen for a collection of bio- 
graphical essays he is working on. 

24. See ends, in WPHC 4, 616/1914; WPHC 4, 1133/1914; WPHC 4, 2376/1914. 

25. These are the dates on their tombstones, which I observed during fieldwork in 1978. 
See also Newell 1906:17. 

26. Dana 1935:189-196; Davis 1892:31; Mahaffy 1909:26; Thurston 1893:8; Samoanische 
Zeitung, 6 Jan. 1912. 

27. Mahaffy 1909:24-26. But on a regional scale the resident trader's day was not quite 
over. Louis Becke (1897:317-319) etched the last act of the drama before the curtain 
finally descended. In doing so he drew the apt comparison with the continuing retreat of 

102 Pacific Studies, Vol. 10, No. 2— March 1987 

James Fenimore Cooper's "gaunt old trapper" Natty Bumppo from the advancing tide of 
civilization in the American West. 

28. See the Tarawa Manager's Annual Reports for 1920 and 1921, BPh; Buckley and 
Klugman 1983:128; Couper 1967:87. 

29. Joseph Mitchell, Notes on the Gilbert, Ellice and Marshall Trades, 22 Nov. 1916, BPh. 

30. Tarawa Manager's Annual Report for 1925, BPh; Buckley and Klugman 1983:137. 


Alefaio, Tanielu 

1979 "Te Okomainga o te Lotu ki Vaitupu," Te Lama (Funafuti), April-June: 16-17. 

Allen, Ernest Tanumafili 

n.d. [Untitled manuscript on Captain E. F. H. Allen and the Samoa Shipping and 
Trading Company, by his son.] Xerox copy provided by Hugh Laracy, in my pos- 

Becke, Louis 

1880 Becke to his mother, 8 July 1880, BP A1372 2 . 
1897 Wild Life in Southern Seas. London: T. Fisher Unwin. 
1905 Notes from My South Sea Log. London: T. Fisher Unwin. 
1909 'Neath Austral Skies. London: John Milne. 

Brady, Ivan 

1975 "Christians, Pagans and Government Men: Culture Change in the Ellice 
Islands." In A Reader in Culture Change, vol. 2, Case Studies, edited by Ivan A. 
Brady and Barry L. Isaacs. New York: Schenkman. 

Bridge, Cyprian 

1883 Report on the Islands of the Ellice, Gilbert (or Kingsmill), Marshall, Caroline 
and Peletu Groups, Visited by H. M.S. "Espicglc" in . . . 1883 (print). RNAS 16; 
WPHC 4, 172/1883. 

Brookfield, H. C. 

1972 Colonialism, Development and Independence: The Case of the Melanesian 
Islands in the South Pacific. Cambridge: Cambridge University Press. 

Buckley, K., and K. Klugman 

1981 The History of Rurns Philp: The Australian Company in the South Pacific. Syd- 
ney: Burns, Philp. 

1983 "The Australian Presence in the Pacific": Rurns Philp, 1914-1946. Sydney: 
George Allen & Unwin Australia. 

Chalmers, Mrs. (translator) 

1872 "Elikana's Story," juvenile Missionary Magazine (1872): 101-105, 123-127, 147- 
150, 175-177, 196-198. 

Chambers, Anne 

1984 Nanumea. Atoll Economy: Social Change in Kiribati and Tuvalu #6. Canberra: 
Development Studies Centre, Australian National University. 

Resident Traders in Tuvalu 103 

Chambers, Keith S. 

1984 "Heirs of Tefolaha: Tradition and Social Organisation in Nanumea, a Polynesian 
Atoll Community." Ph.D. dissertation. University of California, Berkeley. 

Claxton, A. E. 

1889 "Report of Work at the Union, Ellice, and Gilbert Groups, the 'North West Out- 
stations' of the Samoa Mission, L.M.S. Visit of the Deputation 1889." SSR 3/171. 

Couper, A. D. 

1967 "The Island Trade: An Analysis of the Environment and Operation of Seaborne 
Trade among Three Island Groups in the Pacific." Ph.D. thesis, Australian 
National University. 

1968 "Protest Movements and Proto-Cooperatives in the Pacific." Journal of the Poly- 
nesian Society 77 (3): 263-267. 

1973 "Islanders at Sea: Change and the Maritime Economies of the Pacific." In The 
Pacific in Transition: Geographical Perspectives on Adaptation and Change, 
edited by Harold Brookfield. Canberra: Australian National University Press. 

Dana, Julian 

1935 Gods Who Die: The Story of Samoa's Greatest Adventurer. New York: Mac- 

David, Mrs. Edgeworth 

1899 Funafuti, or Three Months on a Coral Atoll: An Unscientific Account of a Scien- 
tific Expedition. London: John Murray. 

Davies, S. H. 

1873 "Report of the Visit to the Ellice and Gilbert Groups in the John Williams 1873." 

SSL 34/2/B. 
1880 "Journal of a Missionary Voyage in the Tokelau, Ellice and Gilbert Groups . . . 
1880." SSJ 176. 

Davis, E. H. M. 

1892 Papers Respecting the Declaration of a British Protectorate over the Gilbert 
Islands by Captain Davis, of Her Majesty's Ship, "Royalist," and General Report 
upon the Gilbert, Ellice, and Marshall Islands. Foreign Office Confidential Print 
6269. London: Foreign Office. See also RNAS 17; CO 225/39/23076. 

Farrell, Andrew (transcriber) 

1928 John Cameron's Odysseys. New York: Macmillan. 

Firth, Stewart 

1978 "Captain Hernsheim: Pacific Venturer, Merchant Prince." In More Pacific 
Islands Portraits, edited by Deryck Scarr. Canberra and Norwalk: Australian 
National University Press. 

Furnas, J. C. 

1951 Voyage to Windward: The Life of Robert Louis Stevenson. Apollo ed. New York: 
William Sloane Associates. 

Gill, William Wyatt 

1872 "Diary of a Tour of the Gilbert, Ellice, Union, and Loyalty Islands in the John 
Williams . . ." ML B1444. 

104 Pacific Studies, Vol. 10, No. 2— March 1987 

Graeffe, Eduard 

1867 "Reisen nach Verschiedenen Inseln der Sudsee," Das Ausland (Augsburg), 
40:1139-44, 1159-64, 1184-91. 

Hayter, Francis 

1871- "Logbook and Journal, H.M.S. Basilisk." PMB 626. 

Hezel, Francis X. 

1983 The First Taint of Civilization: A History of the Caroline and Marshall Islands in 
Pre-Colonial Days, 1521-1885. Pacific Islands Monograph Series No. 1. Hono- 
lulu: University of Hawaii Press. 

Kemble, James Haskell, ed. 

1966 To California and the South Seas: The Diary of Albert G. Osbun, 1849-1851. 
San Marino: The Huntington Library. 

Kennedy, Paul M. 

1974 The Samoan Tangle: A Study in Anglo-German-American Relations, 1878-1900. 
Dublin: Irish University Press. 

LeHunte, G. R. 

1883 "Report of Cruise in H.M.S. 'Espiegle.' " WPHC 4, 157/1883. 

n.d. Six Letters from the Western Pacific. Colombo: Government Printer. 

Macdonald, Barrie 

1982 Cinderellas of the Empire: Towards a History of Kiribati and Tuvalu. Canberra 
and Miami: Australian National University Press. 

Mahaffy, Arthur 

1909 "Report ... on the Gilbert and Ellice Islands Protectorates." CO 225/86/26804. 

Marriott, John 

1883 "Journal of a Visit to the Tokelau, Ellice & Gilbert Groups in the John Williams 
in . . . 1883." SSJ 180. 

Maude, H. E. 

1967 "Louis Becke: The Traders' Historian." Journal of Pacific History 2:225-227. 

1968 Of Islands and Men: Studies in Pacific History. Melbourne: Oxford University 

1981 Slavers in Paradise: The Peruvian Labour Trade in Polynesia, 1862-1864. Can- 
berra: Australian National University Press. 

Maxwell, W. H. 

1881 Report on the Gilbert, Ellise [sic], and Other Islands (print). RNAS 15; CO 225/ 
7/16817; CO 225/42/19817. 

Moresby, John 

1872 "Relative to Skull-Hunting and Kidnapping." British Parliamentary Papers, 

1876 New Guinea and Polynesia . . . London: John Murray. 

Munro, Doug 

1982 "The Lagoon Islands: A History of Tuvalu, 1820-1908." Ph.D. thesis, Macquarie 
University, Sydney. 

Resident Traders in Tuvalu 105 

1985 "On the Lack of English-Speaking Tuvaluans in the Nineteenth Century." In 
Suzanne Romaine et al., Papers in Pidgin and Creole Linguistics, 4:133-141. 
Pacific Linguistics, Series A, No. 72. Canberra: Linguistics Circle of Canberra. 

Munro, Doug, and Niko Besnier 

1985 "The German Plantation at Nukulaelae Atoll." Oral History Association of Aus- 
tralia Journal 7:84-91. 

Murray, A. W. 

1876 Forty Years' Mission Work in Polynesia and New Guinea from 1835 to 1875. Lon- 
don: Nisbet. 

Newell, J. E. 

1885 "Report on the Missionary Voyage to the Tokelau, Ellice and Gilbert Islands." 
SSJ 182. 

1906 "Report of the Visitation to the North West (Ellice & Tokelau) . . . 1906." 
SSL 49. 

Nisbet, Henry 

1875 "Report of the Miss y [sic] Voyage through the Tokelau, Ellice, and Gilbert Groups 
in the 'John Williams' in 1875." SSJ 167. 

Pearson, Bill 

1970 Review of Omai: First Polynesian Ambassador to England, by Thomas Blake 

Clark. Journal of Pacific History 5:240-241. 
1984 Rifled Sanctuaries: Some Views of the Pacific Islands in Pacific Literature to 

1900. Auckland: Auckland University Press/Oxford University Press. 

Phillips, Charles 

1881 "Journal of a Voyage to the Tokelau, Ellice & Gilbert Islands during . . . 1881." 

SSJ 178. 
1884 "Report of Voyage to the North West Outstations of the Samoan Mission . . . 
1884." SSJ 181. 

Powell, Thomas 

1871 [Visit to Tokelau, Ellice, and Gilbert Groups 1871 . ] SSJ 160. 
1879 "Report of the Tenth Annual Missionary Voyage in the John Williams to the 
Ellice and Gilbert Groups . . . 1879." SSJ 175. 

Restieaux, Alfred 

MSa "Recollections of Black Tom, Bully Hayes, D. S. Parker and Jack O'Brien." RP. 

MSb "Recollections of Harry Johnston, trader on Nanumea." RP. 

MSc "George Phillimore Winchcombe." RP. 

MSd "About Louis Becke, Bully Hayes and Others." RP. 

Romilly, Hugh Hastings 

1893 Letters from the Western Pacific and Mashonaland, 1877-1891. London: David 

Rooke, Eustace 

1886 Reports of Commander Eustace Rooke, H.M.S. "Miranda," of Proceedings when 
Visiting the Islands of the Union Group, the Phoenix Group, Sophia and Rotu- 
mah [sic] Islands, the Ellice Group, and the Gilbert Group, April to July, 1886 
(print). RNAS 17; WPHC 4, 84/1886. 

106 Pacific Studies, Vol. 10, No. 2— March 1987 

Shineberg, Dorothy 

1967 They Came for Sandalwood: A Study of the Sandalwood Trade in the South- 
west Pacific, 1835-1860. Melbourne: Melbourne University Press. 

Sollas, W. J. 

1897 "The Legendary History of Funafuti." Nature 1424:353-355. 

Stackpole, Edouard A. 

1953 The Sea-Hunters: The New England Whalemen during Two Centuries, 1635- 
1835. Philadelphia: J. B. Lippencott. 

Stevenson, Robert Louis 

1900 In the South Seas. London: Chatto & Windus. 

Stevenson, Mrs. R. L. 

1914 The Cruise of the "Janet Nichol" among the South Sea Islands. New York: 

Thurston, J. B. 

1893 "Journal Kept . . . during His Cruise to Inaugurate the British Protectorate[s] in 
the Gilbert and Ellice Islands. . . ." National Archives of Fiji (consulted on 
microfilm at the Department of Pacific & SE Asian History, Australian National 
University) . 

Turner, George A. 

1878 "Report of a Missionary Voyage through the Tokelau, Ellice and Gilbert Groups 
in the 'John Williams,' during 1878." PMB 129. 

Waite, Edgar R. 

1899 "The Fishes of Funafuti (Supplement)." The Atoll of Funafuti . . . Australian 
Museum Memoir #3. Sydney: Australian Museum (1896-1900). 

Wallin, F. 

1910 [Report of 30 January 1910 on the Gilbert, Ellice and Marshall Islands.] BPh. 

Werner, B. von 

1889 Ein deutsches Kriegsschift in der Sudsee. Leipzig: Brockhaus (draft translated by 
Ian Grady). 

Wilson, W H. 

1886 [Report of Missionary Voyage to the Tokelau, Ellice and Gilbert Islands in the 
'John Williams; 1886.] SSJ 183. 

Winchcombe, G. P. 

1881- "Notes and Diary." BP. 


Woodford, CM. 

1884 "Journal of a Voyage from Suva, Fiji, to the Gilberts and Back." Woodford 
Papers (microfilm copy in Department of Pacific & SE Asian History, Australian 
National University). 

Young, James Lyle 

1875- "Private Journal." PMB 21. 



Valerio Valeri, Kingship and Sacrifice: Ritual and Society in Ancient 
Hawaii. Translated from the French by Paula Wissing. Chicago: 
University of Chicago Press, 1985. Pp. xxviii, 446, index. $22.50 
paper, $55.00 cloth. 

Review: John Charlot 
East- West Center, Honolulu 

The publication by a prestigious university press of a major book on the 
little-studied subject of Hawaiian religion is an important event for the 
development of a scholarly field. Such a book deserves careful atten- 
tion, especially from the small number of other students of the subject. 
My purpose in this article is to make a detailed examination of Valerio 
Valeri's methods and major conclusions in this book. In so doing, I am 
continuing a work of criticism I began some years ago on the French 
typescript of an earlier version of this work (xv) . 1 A subject as large as 
Hawaiian religion can naturally support a wide variety of opinions, and 
frank and open discussion is necessary for the arduous process of devel- 
oping consensus on points of substance and proper method. 

Many aspects of Valeri's work can be commended. Such a compre- 
hensive book can remind scholars of neglected aspects of the subject. 
Moreover, an impressive number of sources have been used in the book. 
Most important, Valeri, unlike too many scholars, works from the 
Hawaiian language and uses diacritical marks. He also works from 
Hawaiian manuscript materials, which are sometimes quite different 
from the translations made of them (e.g., Malo 1951). In using these 
materials, he often makes good points about proper method. In the lat- 

Pacific Studies, Vol. 10, No. 2— March 1987 


108 Pacific Studies, Vol. 10, No. 2— March 1987 

ter part of the book, he provides useful descriptions of Hawaiian cere- 
monies based on correlations of different sources. 

These positive aspects and many others will be evident to readers of 
Valeri's book. I will now concentrate on some of the criticisms I feel 
must be made of it. 

Valeri states that writing the book constituted for him an "interpre- 
tive experience" of the "dialectical relationship between theory and 
interpretation." The more he "understood the logic of Hawaiian 
thought, the clearer certain crucial anthropological problems became," 
and vice versa (x). His book is best understood, I would argue, through 
an examination of the relation of his theory to the evidence. This is, I 
realize, a pressing issue in anthropology. For instance, Alfonso Ortiz has 
demonstrated that the theories of Claude Levi-Strauss cannot be ap- 
plied to the Tewa. 2 After the publication of two major works on Hawai'i 
with a more empirical orientation (Linnekin 1985; Kirch 1985), it is 
instructive to study an exemplary model of another approach. 

I start, therefore, not with Valeri's views but with his methods, espe- 
cially his relation to his materials: historical documents, texts in Hawai- 
ian and European languages. Valeri's book is interesting among other 
reasons as an example of a transition by some anthropologists from 
fieldwork to historical documentary research. The problems Valeri 
encounters are thus instructive and significant. Valeri makes valid theo- 
retical points about method (xvii, xxiv, xxviii, 66, 96-97, 191-192), but 
criticisms can be made of his practice. (I will not discuss problems of 
translation) . 

A major problem of Valeri's book, in my opinion, is the quantity of 
inaccurate references. I give below a few examples and will add others 
in later sections of this article. 

1. Valeri argues (149) that chiefs must maintain their purity partly 
"through their own comportment." He gives as an example: "divine 
ali'i . . . are obliged — men and women — to remain virgin until mar- 
riage." His source (Pukui, Haertig, and Lee 1978, 1:88-89) refers, how- 
ever, only to women: "Among ranking alii, girls were required to be 
virgin until the first planned union to conceive a child. This was a kind 
of precautionary virginity. Sexual adventure before this royal mating 
could well upset the genealogical applecart!" (Pukui, Haertig, and Lee 
1978, 1:89). This virginity was maintained, therefore, for purely practi- 
cal genealogical reasons, not for the maintenance of ritual purity. More- 
over, the authors go on to say that the emphasis on virginity in some 
Hawaiian legends is a result of missionary influence and that, in reality, 
practice was loose. 3 

Book Review Forum 109 

After a semicolon, Valeri continues, "chastity belts are even used." 
His first reference, which is to the same work (Pukui, Haertig, and Lee 
1978, 1:91), does not, however, discuss virginity before marriage, but 
fidelity within marriage. The example cited by the authors (Fornander 
1916-1917, 4:172-173; see also 164-167) refers to the wife of an impo- 
tent and jealous husband. Moreover, the text belongs to the Mo'ikeha 
stories, which are nontraditional in form and thus arguably nineteenth- 
century compositions. Valeri's other example (Fornander 1916-1917, 
4:112-114) is also a Mo'ikeha story about a woman living with a man 
who is unfaithful to her; for revenge, she binds herself to prevent his 
having intercourse with her. 

Not one of Valeri's references supports his point. Indeed, the thrust of 
his first reference can be used against his view. 

2. Valeri states that there was a Hawaiian "belief" that the ali'i "just 
like the gods, have natural 'bodies' along with their human form. . . . 
this is particularly true of mythical ali'i, who are readily placed at the 
origin of certain species, especially foods" (146). Of the three examples 
he gives, the first (Fornander 1918-1919, 5:266) could better be under- 
stood as a story of the gods; and the second (ibid., 270, 272) does not 
state that the persons involved are chiefs. The third (ibid., 279) fits, but 
one example does not demonstrate a regular "belief." 

Valeri then writes, "As for living ali'i, it is believed that they often 
manifest themselves in powerful and sometimes terrible animals in 
order to punish or protect their subjects." All of Valeri's references are to 
chants in which a chief is called various animals, such as a shark, sting- 
ray, or frigate bird. Such metaphors are common in Hawaiian poetry 
and cannot be used to demonstrate that the chiefs in question actually 
assumed the bodies named. Moreover, the chiefs honored in the chants 
are historic figures: Kalani'opu'u and Kamehameha. Had they been 
able to assume nonhuman bodies, it would have been mentioned in the 
many historical accounts we have of them. (The final reference in 
Valeri's note 40 is to one of the legends he referred to when discussing 
mythical chiefs.) 

Valeri argues against taking such poetic statements as "simple meta- 
phors" (151). The chiefs have a "true affinity" to natural phenomena: 
"Thus, the king is not only compared to the shark, he is the shark 
because he can act through this animal, because he has a substantial 
relation to it, because he is its descendant." I myself have used such 
points to describe the Hawaiian view of the close relationship of all 
human beings to the other elements of the universe (Chariot 1983a: 
e.g., 35-44, 62), but these points do not prove an identity of the chief 

110 Pacific Studies, Vol. 10, No. 2— March 1987 

with the shark or that he can assume a shark body. Valeri himself states 
higher on the same page, "the king is compared to a shark," and has 
himself insisted on the metaphorical character of chant (148) . 

3. Valeri states, "Sexual intercourse with inferiors is also polluting to 
superiors" (91; also 149). This is surprising because, as Valeri himself 
shows (150, 372 n. 56), such intercourse was extremely frequent, 
indeed, a stock theme of Hawaiian literature. The text from Malo that 
Valeri cites in support of his statement refers to marriage with the 
kauwa or pariah class, definitely a special case (Malo 1951, 70-71). 
Moreover, such a marriage is presented as bad for genealogical reasons 
— that is, one becomes declasse — rather than for reasons of pollution. 4 

Valeri's second reference (Kamakau 1961, 128; used also on 149) 
describes Kahahana's losing certain Maui kapu privileges by having 
intercourse with "the lesser chief esses." Chiefly kapu are, however, a 
complicated subject (Chariot 1985, 10-11, 37-40). Kapu often have 
particular rules, and one cannot generalize from one example. For 
instance, another Maui kapu, the Po'oho'olewa i ka la, required its 
owner to shield her head from the sun's rays (Sterling and Summers 
1978, 243). One could not argue from this one example that doing so 
was a general practice; it was simply the regulation of a particular 
kapu. Valeri's references do not, therefore, support his point. 

In his related note (361 n. 12) Valeri states, "Note also that for a high- 
ranking woman the loss of virginity involves a loss of purity and mana." 
His one reference is to the nineteenth-century novel Laieikawai by S. 
N. Hale'ole. Despite its many qualities, the novel cannot be used as a 
reliable guide to classical Hawaiian culture, especially when dealing 
with sexuality and the novel's heroine, who receives in many ways a 
Victorian idealization. 

4. Generalizations can be made only with caution from individual 
authors or works of literature. For instance, chiefs and their lands were 
often connected, but the extreme aspects of identification cited by 
Valeri (146, par. 2; see also 152) can be found only in Ke'aulumoku, 
who, however important, represents a very personal, uncommon view- 

5. Valeri's references to secondary literature also need to be exam- 
ined. He states that chiefs "are characterized by immobility and inactiv- 
ity" (147). That Valeri takes immobility literally can be seen from other 
statements (272, 336). Neither of Valeri's references to G. W. Kahiolo 
and Samuel H. Elbert supports his view; they both portray chiefs as 
delaying a desired action and finally accomplishing it, a common motif 
in Hawaiian literature. To support his view, Valeri has only a pejorative 

Book Review Forum 111 

remark by a foreigner, amply refuted by contemporary literature (147 
andn. 48). 5 

Such examples could be multiplied, and I will mention some in my 
discussion. The above suffice to show that Valeri's references must be 
checked by any serious user of his book. 

Such references often depend, of course, on Valeri's interpretations of 
texts and here also problems can be found. Valeri often announces his 
interpretation of a text rather than offering arguments in support of it. 
For instance, he interprets a farming chant as "a verbal replica of the 
transfer of forces that is the condition of the success of the rite. . . . 
Like the waves coming from over the horizon to break on the Hawaiian 
shores, the gods come from Kahiki to bring life" (55). An examination of 
the chant reveals that this interpretation is based on only one line — a 
reference to a wave from Kahiki in a stereotyped wave list 6 — a small 
basis for understanding a thirty-two-line chant. An examination of the 
rest of the chant can provide an interpretation that accounts for more of 
its lines (Chariot 1983b, 64-65). Moreover, the idea that the gods must 
come from Kahiki every time they are invoked requires support. Valeri 
elsewhere recognizes the existence of the wao akua, "the uplands of the 
gods" (V: "hinterland of gods"), where they can reside (273). The jour- 
ney of Lono from Kahiki to Hawai'i is considered a major event (8). 

Similarly, Valeri wants to use a story to prove that god, sacrificer, and 
victim are identified with each other (132). The story is about a priest, 
named Kanehekili after the thunder god, whose body, when he dies, is 
divided and distributed to people who establish the cult of the thunder 
god in their particular locations. The two problems for Valeri's interpre- 
tation are that the story is about the priest, not the god, and that he 
dies, rather than being sacrificed. Instead of addressing these problems, 
Valeri simply states, "worship of the god is made possible by the victim- 
ization of his priest, who is obviously identified with him (he even bears 
his name)." That is, Valeri takes as given precisely the two points he 
should prove. Alternatives are possible: the body of the priest could be 
efficacious without being considered a sacrifice, and so on. 

On the other hand, Valeri dismisses texts that do not fit his views. 
After developing a theory of mana, Valeri writes, "But in certain con- 
texts the word mana seems to be banalized, to lose its connection with 
both god and community" (100). After using part of a chant, he writes 
(391 n. 81), "I am not considering the entire chant because it is some- 
what anomalous and most difficult to interpret." The chant is no more 
obscure than others Valeri uses. It just does not fit his theory; it is 


Pacific Studies, Vol. 10, No. 2— March 1987 

When Valeri does explicitly analyze texts, other problems in method 
become evident. For instance, he can take one sense of a word and 
ignore others. He notes in the Kumulipo "a curious detail. The god 
Kane bears the name of his worshiper, the human male (kane)." The 
first man, on the other hand, "is called Ki'i ('image'), the generic name 
attributed to the anthropomorphic images of the gods used in worship." 
The result is, Valeri concludes, "a man named as god and a god named 
as man" (6) . A glance at the dictionary will show that Valeri is selecting 
only one sense of ki'i. Moreover, the primary and strict sense of kane is 
"male." 7 The word can be used for the human male, but for any other 
male as well: animal, vegetable, mineral, or god. 

Another problem in method — unfortunately widespread — is using 
too many senses of a word. Polynesians' wordplay with the many hom- 
onyms of their languages is impressive, and the temptation to extend it 
by using the multiple meanings of a word now conveniently provided 
by dictionaries has proved irresistible to most workers in the field (e.g., 
Chariot 1983a, 87, 91). The problem is knowing when to stop. For 
instance, a ritual formula for weakening the god Kamapua'a contains 
repetitions of the word lau. Valeri writes (51), "In my view this is 
because of its double meaning, 'numerous' and 'seine' (PE, 179). Lau is 
a trick word, indeed, for 'numerous' offerings entice Kamapua'a and 
paralyze him in a 'seine.' " A pig in a seine is an unusual image. There 
are in fact no Hawaiian stories or accounts of a pig being caught in one, 
certainly not Kamapua'a. 

Examples of using too many meanings or uses of a word abound in 
Valeri's work. A tapa cloth around the waist of an image is called a 
piko, "navel," and cut in clear imitation of the ceremony for cutting the 
navel cord of a child. Valeri finds it "probable" (316), even though 
unsupported by textual evidence, that the cord represents two other ref- 
erences for piko: genitals and crown of the head or fontanel. A child can 
be circumcised, but how could the crown or fontanel be cut? 

To give another example, Valeri mentions the ritual expression "lele 
wale ka 'aha, 'the 'aha (sacrificial rite, prayer) has flown away (lele)\ 
. . . Lele also means 'messenger'. . . . The altar is indeed a 'messenger' 
that allows men and gods to communicate" (386 n. 20). 8 

Quite large sections of Valeri's book can be based on such arguments 
(e.g., 294-299). Valeri combines a number of traditional uses of the 
place name Kahiki (8-9). The most common is that of a foreign land 
from which gods and other strangers come. Valeri combines this sense 
with references to cosmic points. He has also introduced the word 
"invisible," which does not appear in those texts. The result of this oper- 

Book Review Forum 113 

ation is a definition of Kahiki as "the indeterminate and invisible tran- 
scendent place . . . conceived as that which encompasses the visible 
world." This definition does not fit accounts of travelers sailing to 
Kahiki and setting foot on land. 

Valeri also divides words into parts to get more meanings. Kauila 
wood can yield ka uila, "the lightning": "a manifestation of divine 
power in its luminous but violent . . . form" (269). 

A characteristic method of Valeri's is to systematize texts. Valeri cau- 
tions against this method (191): "each of the principal texts must be con- 
sidered separately and the elementary rules of source criticism be 
applied ... to carefully evaluate the differences between the sources 
and to resist the temptation to arbitrarily construct a single account of 
the rites patched together from different sources." However, Valeri 
writes that two gods who, in one text, "originate in Po, are said in other 
texts to come from Kahiki" (8). This observation is used, among other 
things, to connect Po and Kahiki. Later in the book (331-332) he puts 
together different versions of a story to make his point. Other examples 
will be discussed below. 

Valeri's arguments from texts are often tenuous. In the Kelou Kama- 
kau text on a ceremony (289), the mood during an evening respite is 
described as 'olu'olu, a reduplicative of 'olu, a word with many glosses: 
"cool, refreshing, soft . . . pleasant, comfortable; polite, courteous," 
and so on (Pukui and Elbert 1971, s.v. " o/m"). Valeri uses "affable." 
Later (307-308), Valeri argues that 'olu'olu is "the very word Malo uses 
to describe the effect the sacrifice has upon the god. This highlights the 
correspondence between the state of the god and that of society; thus 
K. Kamakau's text makes it clearer that the god's 'affability' is the result 
of the 'affability' reigning in society rather than the reverse, since the 
latter precedes the former." 'Olu'olu is a common word, so it would be 
difficult to make the correspondence argument even if it were appear- 
ing in the same text. To find the same word in two different texts at dif- 
ferent points of the account of the ceremony and draw a connection 
between them is tenuous indeed. 9 Valeri, however, does not stop there. 
Because the word appeared earlier in reference to the people and later 
in reference to the god, the 'olu'olu of the latter is, he states, the "result" 
of that of the former. That is, not only is there a correspondence 
between the two, but a causal relationship. 

Valeri often draws conclusions such as the above, conclusions that go 
far beyond any support provided by the evidence. Valeri states that 
"dance is necessary to help develop the fetus of an ali'i and to ease his 
birth." He goes on to argue that dance "contributes to affirming the 

114 Pacific Studies, Vol. 10, No. 2— March 1987 

reality of the ali'i's mana." He then goes a step further: "This is why it is 
believed that by dancing people help engender their ali'i" (218). No ref- 
erence is given. Valeri goes even further: "Just as dance engenders the 
infant ali'i, it can be supposed that during the Makahiki it also engen- 
ders the god Lonomakua." A paragraph later, one reads, "The engen- 
dering of Lonomakua, like that of any god . . ." (219). A few pages 
later, the ceremony "presupposes that Lonomakua is rather inoffensive 
at the outset, since he is born of the feasting and thus is already very 
close to humanity" (223; compare 394 n. 146; I will give further exam- 
ples later of Valeri's hypotheses becoming confirmed facts, e.g., 99 and 
101). In a similar fashion Valeri moves from the fact that certain birds 
perch on a type of tree to the statement that the statue made from such a 
tree "is inseparable from the birds" (272-273) . 

Valeri's arguments are often very short. Page 161, paragraph 3, con- 
tains a series of such arguments. One is that the king's rivals "are not 
only enemies, but also close relatives of the victor. Hence they are his 
doubles" (161). Similarly, "the king's human sacrifice is always a fratri- 
cide: either a literal one ... or a metaphorical one — since every trans- 
gressor implicitly identifies with him and therefore becomes his 'dou- 
ble' " (165). And "Atea, who in central Polynesia is a symbol of clarity 
and light and as such is a double for Kane" (169). For a final example, 
"Kahoali'i not only is a human incarnation of Ku but is also, by the 
same token, the king's double, as witnessed by his being called the 'royal 
companion' " (325). Many arguments would be necessary to support 
such conclusions. 

There are a large number of such short arguments in the book. "The 
tooth sacrificed is probably a substitute for the whole person, since to 
dream of losing a tooth means death" (355 n. 27). "While a priest is 
announced by a single rainbow, a high-ranking ali'i is announced by 
two rainbows. . . . In short, one ali'i equals two priests" (369 n. 19). 

Valeri does not hesitate to introduce his speculations into an interpre- 
tation: "The mode of the pig's death indicates that it is a piglet, since 
otherwise it could hardly be lifted by one person and dashed against the 
ground. It thus seems that by incorporating a very young victim in Ku, 
the god's rebirth in a new form (Kunuiakea) is brought about" (313). I 
will not discuss the validity of his argument or conclusion, except to say 
it is important for the theme of his book. I will note, however, that 
Hawaiian pigs were in fact smaller than modern Western ones and, 
even full grown, could have been handled as described by a robust 
man. 10 

Very often, Valeri eschews argument, covering a conclusion with a 

Book Review Forum 115 

word like "clear," "evident," "obvious," or "likely." Some longer phrases 
are "It is immediately clear, then . . ." (7); "It must follow . . ." (58); 
"We must deduce . . ." (86); "we must suppose . . ." (86, 111-112); "it 
seems difficult to deny . . ." (86); "seems to indicate . . ." (86); "We 
may conclude, then . . ." (87); "It seems reasonable to suppose, then 
. . ." (87); "It seems to me that this ritual clearly displays . . ." (87); 
"the evidence suggests that, at least to some extent . . ." (88); "legiti- 
mate to assume . . ." (98); "Putting all these clues together, I feel 
inclined to hypothesize that . . ." (99); "one is tempted to define . . ." 
(133); "It is difficult not to recognize in these . . ." (251); "The text by 
Wilkes prompts another reflection . . ." (308); "one cannot help relat- 
ing . . ." (316); "Is it too audacious to suppose . . ." (326); "Clearly 
what is implied . . ." (391 n. 92). Such phrases create much of the 
impression of the book. 

The hypothetical character of his points does not prevent Valeri from 
using them as if they were confirmed. This can happen very quickly. 
"This classification of the fish species is in large part hypothetical. It 
does in any case confirm the theory advanced . . ." (26) . "Although this 
last proposition is speculative, it is confirmed by the fact that . . ." 
(325). We read that the feather god referred to could be "Kuka'ilimoku 
or some equivalent god" (222); then the very next sentence states, "the 
fact that Kuka'ilimoku obtains a place . . ." At the end of the para- 
graph, we read of the imposition "of the god of force (Kuka'ilimoku) 
over the god of the festival (Lonomakua)." 11 As a result of such methods, 
Valeri's argumentation often seems nothing more than strings of hy- 
potheses, each being treated as a fact by the one that follows (e.g., 315- 
317). These freshly minted hypotheses are so firm in Valeri's mind that 
he can berate the writer of an earlier book for not taking one into 
account. 12 

Valeri's entire book rests in fact on a circular argument: his theory 
developed in the first part of the book will be confirmed by his analysis 
of the ritual in the second (e.g., 191). But that analysis depends on his 
theory. For instance, he writes about his view of the hierarchy of gods: 
"the scheme I have presented is a simplified model; nevertheless I 
believe it is not an arbitrary construction. On the contrary, it will be 
confirmed by the analysis of the temple system" and the ritual (110). 
However, in his chapter on "The Hierarchy of Temples," he finds so 
many problems and must use so many qualifications (184-187), that he 
is forced to conclude: "The details I have just enumerated are so many 
limits on the validity of the model. . . . Nevertheless, I believe this 
model offers a valid presentation of the ideal background justifying the 

116 Pacific Studies, Vol. 10, No. 2— March 1987 

concrete temple hierarchy. At any rate, on this point as on many others 
we are reduced to speculation, for the actual relations between individ- 
ual temples are very poorly known to us" (187). He has, however, pre- 
sented enough evidence to cast the very idea of such a hierarchy in 
doubt. In any case, since neither passage gets beyond a simplified and 
speculative model, neither can confirm the other. 13 

Valeri's attitude toward his evidence can be apprehended in an 
important passage (96-98). Wanting to elucidate the meaning of mana, 
he mentions that "Firth studies the various meanings of mana as they 
occur in all texts he recorded." "As a starting point, this method is of 
course a must." He concludes, "we must study the textual occurrences of 
mana in their contexts." Unfortunately, "the word mana does not fre- 
quently occur in the texts." After looking at the few examples, he writes: 
"However, it would be wrong to make much of the rare occurrence of 
mana in these descriptions. For one thing, Malo and Kelou Kamakau 
give only very few of the prayers that were uttered in the ritual, in 
which the word mana must have been included rather often." Valeri's 
argument for this point is that all the useful examples he finds in K. Ka- 
makau do occur in prayers. But he has shown that those examples are 
very few, and he has proper reservations about accepting others. 

On what then does he base his view that the word "must have been 
included rather often"? In a similar passage (145), he writes: "even if 
the ancient texts in which the word akua refers to ali'i are few, it can be 
safely assumed that this usage really existed because it follows necessar- 
ily from the attribution to ali'i of the fundamental properties of the 
divine." Valeri is deducing evidence from his theory. 

Valeri continues arguing against finding the rarity of mana "exces- 
sively relevant." The rituals "clearly involve the transmission of mana"; 
the texts are not "complete accounts" of the rituals "and especially of 
what is presupposed throughout it and therefore does not have to be 
explicitly stated." Though Malo and K. Kamakau "do not explicitly 
say so," another source— which Valeri argues should reflect Malo's opin- 
ion—does make the point. After further such arguments, Valeri con- 
cludes, "This example shows, I believe, the dangers of a blind literal- 
ism and of the assumption that only verbal statements are informa- 
tive" (98). 

No scholar would want to fall into such errors. Nevertheless, Valeri 
himself states that verbal expressions are clearest (e.g., 343-344), and 
the large quantity of available Hawaiian literature provides unusual 
opportunities for accurate understanding; but it also imposes responsi- 
bilities of precise interpretation and full documentation, in addition to 

Book Review Forum 117 

the fundamental responsibility of not going beyond the evidence. It is 
remarkable that at key points in his work, Valeri must admit that his 
views are not supported by the texts. 14 

Valeri's main thesis can fairly be said to depend on his interpretation 
of one section of the main temple ceremony: "the god 'is born' as a man, 
as the ideal man made possible by an ordered society. . . . The transfor- 
mation of the god into the perfect type of the human male is thus com- 
pleted. Therefore the true human nature of the god becomes fully 
apparent . . ." and so on (314-315; compare 250, 330, 345). Yet an 
examination of the text in question (Malo n.d., 175-176, sections 99- 
101; Malo 1951, 173; Valeri 1985, 309) reveals that the two points Valeri 
is making— that the god is born and that it is "the ideal man" or "the 
perfect type of human male" — cannot be found. 

Valeri puts "is born" in quotation marks (see also 287), but the 
Hawaiian equivalents do not appear in the Hawaiian text. Elsewhere, 
Valeri speaks of the birth as indeed taking place (e.g., 306, 330). What 
then do the quotation marks mean? 

The ceremony described in the Hawaiian text is in fact that of cutting 
the navel cord, as Valeri himself states. Malo's description of the chiefly 
ceremony for male infants (Malo n.d., 141-142; Malo 1951, 136-137) 
shows that that ceremony could be separated from the birth. This is 
even clearer in the text of Kelou Kamakau: 

a puka mai la iwaho, kaawale ae la ia, lavoe ia aku la imua o ke 
alo o ke akua, a me ke alo o na kahuna, hoali ae la ke kahuna i 
ka ohe e oki ai o ka piko 

"and when [the child] came out, he was separate/separated, he 
was taken out before the face of the god and the face of the 
priests. The priest waved the bamboo with which the navel 
cord was to be cut." (Fornander 1919-1920, 6:5, 7) 

The ceremony is then briefly described. 

Valeri has earlier treated a text on the chief 'Umi as a parallel: Liloa 
has 'Umi brought to the temple, "has his umbilical cord symbolically 
cut, and has him undergo other rites, following which he is 'reborn' as a 
noble" (277-278). In neither reference does the word "reborn" appear, 
although Valeri gives it in quotation marks. Moreover both translations 
have the expression 'oki ka piko refer to circumcision rather than to cut- 
ting the navel cord, although the Fornander translation uses both. Fur- 
thermore, the navel cord cutting ceremony can be used as an image sep- 

118 Pacific Studies, Vol. 10, No. 2— March 1987 

arate from birth, as can be seen from a vivid and unusual section of 
Ke'aulumoku's Haui Ka Lani (Fornander 1919-1920, 6:394-395, 11. 

Polynesians could use a number of images to express the beginning of 
something, and the exact extension of an image must be carefully delin- 
eated in any one use. Moreover, the primary object of the ceremony 
under discussion is the statue, and some discussion is necessary on which 
ceremonial points apply to it and which to the god, however one con- 
ceives of the relation between the two. I would argue, therefore, that 
Valeri goes beyond the evidence in describing the section of the cere- 
mony as the birth of the god. His characterization of that god as a man 
is derived wholly from his theory and has no basis whatsoever in the 

There is, moreover, one further difficulty for Valeri's theory. Of the 
three descriptions of the sequence of temple rituals, "Malo is the only 
author who describes the rite of the god's birth" (315). That is, what 
should be the most important ceremony of the whole sequence is 
replaced by a different one in two of the three sources. Valeri admits 
that these "texts differ because they reflect alternative practices rather 
than because one of them reflects the 'true' form of the ritual while the 
others do not" (317). Valeri exerts his considerable powers of argumen- 
tation (315-317) to demonstrate that "the comparison of the different 
versions of the rite reveals new aspects of its meaning and confirms the 
cogency of our previous interpretations. . . . [All three] descriptions 
presuppose the same sense relationships, thus the same 'grammar' ' 
(317). The birth of the god cannot, however, be found in the other 
descriptions. Valeri writes, "The equivalence of the makVilohelohe rite 
in K. Kamakau's account and of the image's birth rite in Malo's account 
appears to be truly paradigmatic because the two rites occur in the same 
syntagmatic position" (317). Again, the element missing is any hint of 
birth. Valeri later passes over this difficulty (334). 

Such difficulties with the textual evidence are not uncommon in 
Valeri's section on ritual. 15 In a characteristic passage he writes, 

The purpose of the rite is then clear; it brings about the growth 
of a god who, having just barely entered the men's temple, is 
like a small infant, still on the threshold of the human world 
into which he must be integrated. This "baby" is in fact not yet 
truly born, since the rite of the god's birth will take place later. 
But we must not take the metaphors of the rite too literally; this 
birth will not be the god's first (since each sacrifice uses the 

Book Review Forum 119 

birth/death metaphor), but only his final and definitive "birth" 
— a little like that of 'Umi in the temple where his father sends 
him to "be born" a second time. (305-306) 

In sum, one can apply to Valeri his criticisms of others: "They appear to 
be rationalizations; it is not clear whether they are produced by the 
informants or by the authors" (51; cf. 261). 

Valeri's attitude toward evidence is best understood, I would argue, 
from his theoretical orientation; he states that his "analysis ... is of the 
structural kind, tempered, however, by as much skepticism and good 
sense as I am capable of" (193; see also 388 n. 37). As seen earlier, his 
book is a result of the "dialectical relationship between theory and 
interpretation" (x). He himself provides the necessary information on 
his philosophical orientation and the sources of his main ideas: "I take as 
my starting point the Hegelian idea that religion is 'objective spirit,' that 
is, the objectified system of ideas of a community" (x). Combined with 
Platonic ideas, this view leads to the theory of Hawaiian gods as "types" 
(31-33): "this idea is personified, given a concrete (albeit imaginary) 
form; therefore it becomes a type" (ix); "the god is a concept, a type" 
(103; see also 74, 100-101, 347, and many other places). Such gods are 
both concrete and abstract (32 and 351 n. 32 for Hegel and Feuerbach 
references) . 

Valeri's interpretation of ritual and sacrifice is also inspired by these 
sources: "the major influence on my thought has been Hegel's Pheno- 
menology, for I have attempted to view Hawaiian ritual as a manifesta- 
tion of the dialectics of consciousness" (xi) . Thus ritual is interpreted as 
reproducing a thought process (e.g., 70-74, 300-301, 308, 323-325, 
345-346, 348). Another influence on this point is Durkheim's reduc- 
tionist theory of "the efficacy of ritual as due to the power of collective 
consciousness on concrete social agents and relations" (xi). 16 

Also from Feuerbach Valeri takes the ideas that "Hawaiian religion is 
essentially anthropomorphic" (xi), that the state is a projection, so to 
speak, of human nature or essence, and that the head of state is "the 
representative of universal Man" (130). I will discuss below Valeri's 
attempt to universalize the anthropomorphic element of Hawaiian reli- 
gion. The above Feuerbachian idea, combined with "the famous Brah- 
manic saying: 'Sacrifice is man. . . . Thus the sacrifice is the man' ' 
(358 n. 65), provides also another central idea of Valeri's interpretation 
of ritual (49, 70-73, 355 n. 28; cf . 64) and of Hawaiian sculpture (248- 

Valeri thus approaches Hawaiian religion with considerable philo- 

120 Pacific Studies, Vol. 10, No. 2— March 1987 

sophical baggage. His writing is in fact remarkable for the number and 
frequent use of terms from Western philosophy and religion, terms 
introduced without explicit justification and used as analytical tools (I 
enclose in quotation marks those so used by Valeri): particular and uni- 
versal (270); "essence" and "matter" (56); essence (138: "conformity 
with the idea of human essence"; 357, 366 n. 24, etc.); "substantialist" 
and divine substance (363 n. 3; used to translate Hawaiian); consub- 
stantial (139); transcendental and non-empirical, immaterial (268, 
etc.); invisible, the divine (discussed below); "supernatural" (350 n. 3); 
"sacred" and "profane" (113); sacred and profane (120); creation (75, 
89, 156, and throughout); incarnates (134); "miracles" (362 n. 26); eter- 
nal life and sacrificial death (328); Utopia (226). The word "mandate" 
(370 n. 33) is taken from Chinese religio-political thought. 

Moreover, Valeri expresses, again without argument, classic Western 
views or reads them into Hawaiian thinking: on humans being separate 
from nature, on the separation of religion from natural science (35), on 
the nature of women, on sight and intelligence being "what is most 
human" (252), and on the male standing "for the entire human species 
because he is its superior form" (270) . 17 

From his sources and statements it is clear that Valeri represents an 
extremely intellectualist Western philosophical tradition. He speaks 
approvingly of "evoking immateriality in materiality, abstraction and 
generality in the concrete and the individual" (268); of "the passage 
from a particular to a universal and from something natural to some- 
thing human" (270); and states that "the transition from knowledge of 
the visible (which is particular and limited) to knowledge of the invisi- 
ble (which is general and unlimited) can be obtained only by negating 
man's empirical vision. . . . the transformation of man's consciousness, 
which moves from empirical vision to intellectual vision, from the parti- 
cularity of percept to the universality of concept" (324; see also 325). 

Valeri's extreme intellectualism can be seen in his applying uses from 
logic of terms like "identical" and "substitutable" (93) to things that are 
not merely ideas. In doing so, he is following or extending structuralist 
practice. My purpose in this article is, however, not to discuss struc- 
turalism itself, but only Valeri's application of it to Hawaiian religion 
and culture. In Valeri's perspective, the great enemy chiefs Keoua and 
Kamehameha can be described as "perfectly identical" (162); 18 Palila is 
"interchangeable with the god he worships" (276); "the king and his 
adversary form a pair of absolutely identical terms" (279); "the in- 
terchangeability of victim, sacrificer, and sacrifier" (308; see also 
389 n. 63). 

Valeri goes even further. In a Hawaiian story a father praises a 

Book Review Forum 121 

heterosexual couple, using a list of traditional expressions of praise for 
people of great physical beauty. Valeri interprets this speech as describ- 
ing the couple as "perfectly identical" (166). But in the speech, there is 
no hint of such a "perfect identity," which would be naturally impossi- 
ble in such a couple. 19 

Valeri goes, however, even further than this: "There is, however, a 
logical and factual connection between the idea of royalty and that of 
twinship. The perfection innate in royalty implies that any plurality 
manifested within it is a plurality of identical beings. In fact, by defini- 
tion, two or several perfect beings cannot be different without some of 
them being imperfect, since their perfection consists in fully instantiat- 
ing the same type" (374 n. 87). In other words, perfection consists in 
looking a certain way, so two perfect beings would have to look exactly 
alike. Not many Western idealists would go to such an extreme. Valeri's 
argument turns on his definitions of "perfection," "type," and "instanti- 
ate," which he imposes without argument on Hawaiian culture. 

Valeri's philosophical bent has a strong influence on his study of 
Hawaiian religion. A characteristic example is his discussion of chiefs' 
having "the fundamental properties of the divine" (145; also 145-153), 
several points of which I have discussed before. Valeri has argued "that 
the completeness, and therefore purity, of the gods is absolute and 
depends on no one but the gods themselves. Indeed a number of things 
considered akua evoke total autonomy and independence," such as the 
watch, compass, and clock owned by Captain Cook and the full moon. 
"For the Hawaiians the circle evokes a being closed in on itself because 
it is complete and self-sufficient." Valeri gives characteristic arguments 
for this, but then admits that "a different notion of the divine perfection 
is suggested by the Kumulipo." He concludes, "two notions of divine 
perfection or purity seem to coexist" (88-89) . I will not discuss this sec- 
tion except to remark that many nonround objects were considered 
akua and that a traditional expression for feminine beauty — mahina ke 
alo, "the face is like a moon" — does not evoke autonomy and self-suffi- 

Turning to chiefs, Valeri finds them like gods in "their association 
with emblems of transcendence and perfection. . . . celestial entities 
. . . whose circular form and movements connote perfection" (146). 
Valeri then concludes, "Conceived of as autonomous, from one point of 
view the ali'i are also thought to be free of desire, precisely like the 
gods" (147). 20 As Valeri himself admits (150), this statement contradicts 
the evidence for chiefs, whose very genitals were lauded publicly in 
chant. 21 

It also contradicts the evidence for gods, which Valeri does not dis- 

122 Pacific Studies, Vol. 10, No. 2— March 1987 

cuss. But he himself has shown that they eat and can be attracted in 
prayer by being offered food (56, 133). Moreover the gods have sex. Pele 
and Kamapua'a are famous for their love affairs. Lonoikamakahiki is 
married in the myth that forms the foundation of the ritual Valeri is 
studying (214-215). KG can be married in other stories. In a Hawaiian 
story, "Komo ihola ka iini iloko o Kane a me Kanaloa no keia ui nohea " 
("Desire entered into Kane and Kanaloa for this handsome young 
beauty," my translation; Green and Pukui 1936, 114). Valeri himself 
says that fertility is divine (273) . 

After discussing the immobility (see above, p. 110) and invisibility of 
chiefs, Valeri writes, "Because he is supposed to be self-sufficient, with- 
out desire or sadness, always in control of himself, a divine ali'i cannot 
openly display his emotions; this is why he expresses them metaphori- 
cally, poetically" (148). Curiously, the one chant to which he refers con- 
tains not a single metaphor — perhaps the only one in all Hawaiian liter- 
ature not to do so (Hawaiian text in Dibble 1838, 95; Remy 1862, 
202-205). All classes of Hawaiian society used poetry, and all the 
poems, except the perhaps unique exception Valeri cites, have meta- 
phors. Moreover, those metaphors, as Valeri argues (151), were not 
"simple" but truly expressive of man's relation to nature. Poetry was 
used in many forms of communication, from joking, to courting, to 
praying, to politicking (Chariot 1985). Moreover, chiefs could use other 
genres. Stories picture them conversing, telling stories, and so on. Valeri 
shows them giving judgments in the ritual (e.g., 263, 289). I know of no 
evidence that Hawaiian chiefs and chiefesses had rank-related difficulty 
verbalizing. Similarly, there is no evidence that "high-ranking nobles 
always avoid situations where there is laughter, since laughing puts an 
end to taboos" (287). Kamehameha is reported making his courtiers 

I cannot imagine a description of Hawaiian chiefs further from the 
richly available evidence. The arhat-like ideal Valeri extrapolates from 
his interpretation of circles contradicts even his own view of the interde- 
pendence of chiefs and the other elements of society (e.g., 7). Valeri sees 
that his picture contradicts the evidence. His solution — as in the case of 
the gods, discussed above— is to posit "an insurmountable contradic- 
tion" in Hawaiian culture on this point (149-150). The contradiction is 
between his theory and the evidence. 

I would now like to criticize in less detail some of Valeri's main 
points. In accordance with the ideas of Hegel, Valeri announces as a 
major theme that there is a Hawaiian religious "system" (x) just as there 

Book Review Forum 123 

is a system of rituals (189, an idea he seems to get from Mauss) and of 
society ("society as a consistent system," 187). In this system can be 
found "the logic of Hawaiian thought" (x; also, e.g., 192). Valeri uses 
that underlying logic for his arguments 22 and for his reconciliation of 
conflicting sources (192). Because of this logic, which he has "under- 
stood," he is "giving a coherent interpretation of Hawaiian religious 
ideas (the first one, to my knowledge)" (x). The last paragraph of the 
book — a single sentence — seems to refer to the logic of the system as he 
has explicated it: "How could it be otherwise?" 

Valeri writes most often as if there were one historical system that, 
with a little more surviving evidence, could be reconstructed down to 
fine details (e.g., 14-17, 25, 109-128). He will find a system even 
through the welter of inadequate or conflicting evidence (25, 46 par. 3, 
57 par. 4, 186 par. 2). He will sketch out a systematic analysis even if he 
has to admit that it cannot fit all the evidence (102-103, 180-181). He 
says he will avoid systematizing the conflicting descriptions of the ritual 
in his sources, but does. 23 

The evidence of religious disunity is, however, overwhelming — even 
on the testimony of Valeri's book. 24 To accomodate his theories, he him- 
self must posit fundamental contradictions and differences in Hawaiian 

Valeri attempts to solve aspects of disunity (45 par. 4, 225 and n. 62, 
343) . For instance, elements that do not fit his reconstructed system can 
be called "marginal," such as priestesses and prophets (112, 138-139). 
Yet they do take part in the temple ritual (328). Sorcerers, the very 
antithesis of Valeri's system — "in clear opposition to the priests of the 
'central' cult" — and the target of his polemic are "marginal and resid- 
ual" (138). 25 Yet, surprisingly, they are found working alongside the 
priests at court (183, 185, 247-248, 380 n. 9). 

The very idea of a single, unified system is, however, historically 
implausible. Hawai'i was settled around a.d. 500, if not earlier, and 
had an increasing population scattered widely over the large archipel- 
ago. No religious system could have remained static in such circum- 
stances. Indeed, Valeri's system, as he himself states, is an "ideology 
imposed by the aristocracy" (19; also 348), the product of "a powerful 
class of priests, that is, of professional intellectuals. This systemic, 
priestly view of the pantheon . . ." (36). That priestly group and the 
high chiefly class with which it worked were, however, late develop- 
ments in Hawaiian culture (e.g., Kirch 1984, 257-262; Kirch 1985, 
e.g., 301-308). Their religion presupposes then earlier stages or forms of 
Hawaiian religion. Valeri himself speaks of "a tendency to unify the 

124 Pacific Studies, Vol. 10, No. 2— March 1987 

pantheon and the entire cult under the major gods" (110). There must 
then have been an earlier, pre-tendency period in which such unity did 
not exist. Would not a genetic approach, the study of the historical 
development of that priestly system, be helpful in understanding it? 
Should Valeri not have put his subject in its historical context? Evidence 
can indeed be found in Valeri's book for the view that elements of that 
system were developed from earlier religious views and practices (126, 
173, 280, 302, 357 n. 56, 384 n. 71, 396 n. 178). For instance, the navel 
cord-cutting ceremony used in the temple ritual is arguably based on 
one used for male chiefly babies, a ceremony that is itself in all likeli- 
hood based on earlier practices surrounding the cutting of the cord 
(navel cords did after all have to be cut). But Valeri proceeds to describe 
the system ahistorically in a sort of "anthropological present." 

In my view, given the evidence for the rich variety of Hawaiian reli- 
gion, no one system could be expected to absorb all the earlier religious 
elements, as will be seen in my discussion below of Valeri's treatment of 
the gods. On the other hand, any system attempting to be comprehen- 
sive would need to absorb such a large amount of material that com- 
plete logical consistency could not be expected. 

In any case, given a population of some two hundred thousand on 
separate islands under independent chiefs, there need not have been just 
one high-priestly system. Valeri in fact admits that "some traces of a dif- 
ferent system exist, especially on the island of Kaua'i" (185; also 335). 
The system Valeri studies is, therefore, "valid mostly for the island of 
Hawai'i" (184). He can be even more specific: he will use Malo's calen- 
dar, "which was the one used in the western part of the island of 
Hawai'i, because the main descriptions of the ritual cycle refer to this 
area" (198). 

None of these considerations prevent Valeri from claiming, "In the 
preceding two parts I have attempted to give a coherent picture of the 
Hawaiian ideological system by considering all available information" 
(191). But to take one geographically, historically, and socially delim- 
ited religious system out of several and call it the system of a culture is as 
incorrect as to claim that the Memphis theology represents all of ancient 
Egyptian religion or the Midewiwin all of Ojibwa. 26 Ake Hultkranz's 
idea of different "configurations" of religious views and practices being 
used in a single society, which he applies to American Indian religions, 
would be a useful tool of analysis. 

Valeri's discussion of the "pantheon" shows that (1) he wants to make 
it all-encompassing for Hawaiian religion as a whole, and (2) that he 
wants it to be coherent. He bases himself, as others have before him, on 

Book Review Forum 125 

systematizing nineteenth-century sources, especially Kamakau, and on 
the conventional idea of "the four great gods of Polynesia": "There is no 
doubt that Ku, Lono, Kane, and Kanaloa are the highest gods" (109). 
As is often necessary for Valeri, he leaves out a good deal of evidence: 
the Pele literature in which those gods are mercilessly subordinated to 
her (e.g., Pukui and Korn 1973, 55; Chariot 1983a, 24), the story of 
Kamapua'a's defeat of Lonoka'eho (Kahiolo 1978, 32-43 and parallels), 
and the many other stories of the conflicts of the gods (Chariot 1983a, 

Curiously, Valeri resurrects notions now recognized as nineteenth- 
century, biblically influenced attempts to rationalize classical Hawaiian 
beliefs: ke koko'oha o ke akua, "the quaternity of the god" (V.: "The 
association of four gods"), an idea based on the Christian Trinity (13). 27 
He also adopts the Trinitarian notions of other nineteenth-century 
Hawaiian writers, who degrade Kanaloa to a sort of demon: "the qua- 
dripartition of the gods is a superficial phenomenon that conceals a tri- 
partition on a deeper level" (18). 28 

A good deal of Valeri's book is spent trying to subsume all the Hawai- 
ian gods under the four he regards as principal (e.g., 13-30). He 
describes the process: "Deities are also created spontaneously and unsys- 
tematically at the lower levels of the pantheon, where their prolifera- 
tion is allowed by the very system I have delineated. Spontaneous cre- 
ation and systematic ordering are thus two dialectical moments in the 
constitution of the divine in this hierarchical society" (36) . As a result of 
this process, "Lower-level deities are either particularizations of these 
gods or personifications of some of their predicates" (109). Valeri con- 
cludes, "All in all, I think that it is possible to view all inferior gods as 
encompassed by the major ones" (110). 29 

Such priestly connection of "lower-level" gods to "higher-level" ones 
is found in Hawaiian religion as it is in Hinduism and Buddhism. But 
again Valeri has taken the most extreme position in claiming that all 
Hawaiian gods were thus connected. His discussion has not demon- 
strated that all family gods and wandering spirits have been so treated. 
Most important, it is impossible to absorb the female gods of Hawai'i 
into the four male gods. Valeri, therefore, belittles them, stating, "god- 
desses are few and have a marginal position in the Hawaiian pantheon. 
This corresponds to the marginal position of women in the ritual sys- 
tem" (19; cf. 12). Goddesses are in fact numerous and important. When 
Valeri does not ignore the great volcano goddess Pele, he groups her 
with "the female deities of sorcery" (112). 30 Sorcerers, he claims, are 
"marginal," yet Pele's priestesses take part in the temple ritual (328, 

126 Pacific Studies, Vol. 10, No. 2— March 1987 

401-402 n. 251). Valeri is careful to state for his source that "these god- 
desses are ultimately controlled by the king" (402 n. 251), a judgment 
that contradicts the Pele chants mentioned above in which supremacy is 
claimed for her. Pele is indeed one of a number of other gods— besides 
Valeri's four principal ones— who take part in the ritual (e.g., 264, 290), 
an indication that they had not lost their identity and utility. Valeri is 
also anxious to deny Kamakau's statement that statues of goddesses were 
placed in the temple (238, 245) . 

From his view of a coherent, all-encompassing hierarchy of gods, 
Valeri draws important conclusions for the religious life of Hawaiians. 
For instance, he holds that an individual's relation to the gods is 
mediated by the hierarchy just as lower-level sacrifices are by higher- 
level ones (e.g., 19, 126, 185). As a result, "the ali'i are continually or 
almost continually in relation with the gods, while commoners are in 
relation with them only occasionally (during holidays, precisely) and 
always by means of a more or less direct mediation on the part of the 
ali'i" (127); "Direct contact with the most important gods of the society 
is possible only for the king and his chaplains" (140). 31 These views are 
contradicted by numerous accounts of visions; family, fishing, and 
farming gods; prayers on many occasions; the marriage of Ku into a 
commoner family (Green and Pukui 1936, 127), and so on. 

The second major principle of Valeri's book — influenced by Feuer- 
bach — is that "Hawaiian religion is essentially anthropomorphic. All 
gods have in common what all subjects have in common: the fact of 
belonging to one single species, the human species" (xi; also, e.g., x, 
272-273, and the argument discussed below). Valeri's exclusive equa- 
tion of "subject" with "human" is unusual. Some worldviews recognize 
nonhuman subjects, such as angels and leprechauns. Valeri's use of 
"subject" corresponds to his use of the phrase "personal gods" as the 
equivalent of anthropomorphic ones: "personal, anthropomorphic gods 
such as Kane and Kanaloa" (7; see also 6, 10, 65). However, in religious 
studies, the phrase "personal gods" can be and often is used of nonan- 
thropomorphic ones. 

Valeri's position seems to be derived more from his theory than from 
evidence. His one argument is that all Hawaiian gods have a human 
body in their kino lau, their system of multiple bodies (9-12, 21, 31, 35, 
47) : "the 'genus' of all species included in one god belongs not to the nat- 
ural world but to the human, social world" (11); "the human form is the 
most generic component, while their natural forms differentiate them" 
(21); "each deity is characterized by two kinds of 'bodies,' . . . : natural 

Book Review Forum 127 

bodies and the human body. . . . the human species is the common ele- 
ment underlying all natural manifestations of the divine. ... all gods 
equally represent the human species" (31). He concludes, "The unity of 
the divine is the unity of the human species" (35). 

This position can be criticized on several grounds. First, the lowest 
common denominator need not reveal the genus or prove that the 
human form is primary. For instance, in Native American religion, ani- 
mal gods can appear in human form, but are still thought to be animals. 
The Buffalo Maiden of the Sioux appeared to two youths as a young 
woman, but as she moved away, they saw her changing back into her 
buffalo body. Animal gods can appear in human form just because they 
are appearing to human beings, not because they are really human. The 
common denominator reveals the nature of the audience, not of the 

Most important, to prove his point, Valeri would have to demonstrate 
that all Hawaiian gods have human bodies. This he does not even begin 
to do (47). 32 On the contrary, in one of the major faults of his book, he 
simply ignores the large number of gods that have only animal or ele- 
mental bodies. 33 

Moreover, the animal body of the god can be presented as primary 
even when he or she has a human body as well (e.g., Green 1926, 64-65 
[rock]). The shark-man is really a shark. Pele tells her attracted sisters 
that Kamapua'a is really a pig. 34 Kahiolo is definite (58-59): "Afca, ua 
pololei no o Pele malaila, no ka mea y he puaa io no oia, 'But Pele was 
right, because he was a real hog.' " A literary motif found in numerous 
stories is that of the marriage between a human being and someone who 
is discovered to belong really to another species. 35 Such gods and stories 
can be found elsewhere in Polynesia. 

Animal gods are a common phenomenon worldwide, and a heavy 
burden of proof lies on anyone who would deny their existence in a 
Pacific Island culture. In so dealing with Hawaiian culture, in which 
animal gods were and continue to be important, Valeri's criticism of 
another scholar for "leaving out of the field ... all that does not fit the 
theory" (66) turns against himself. 36 

The reason for Valeri's strong anthropomorphization of Hawaiian 
religion — other than his intellectual sources — is his presupposition of a 
separation of human beings from "nature": "natural phenomena extra- 
neous to man" (30). This very Western view is used in a Western way: 
"Having become totally dehumanized, nature becomes totally distant 
from man. . . . the humanization of nature is the necessary correlate of 

128 Pacific Studies, Vol. 10, No. 2— March 1987 

its appropriation by man" (78). 37 Thus the ritual is used for the purpose 
of anthropomorphizing nature (72-73, 269-270, 345-346, 353), as are 
other Hawaiian practices. 38 

This view is diametrically opposed to that of the Kumulipo, in which 
human beings are placed on the same family tree with the rest of 
nature. It opposes Valeri's own statements about Hawaiian metaphors 
(151), as seen above, as well as a number of other Hawaiian practices 
(e.g., Chariot 1979). I would argue that Valeri's presupposition of a 
fundamental separation of human beings from nature cannot be found 
in Hawaiian culture. In fact, in the section in which Valeri admits that 
his theory cannot be found in the Hawaiian texts, he states that Hawai- 
ian thinking "presupposes the process of consciousness I am referring to, 
since it presupposes that the world of nature and the world of man are 
comparable and therefore that nature is already humanized and man 
already naturalized" (34) . But Hawaiians need not have gone through 
that "process of consciousness" — even unconsciously, as Valeri seems to 
say — and need not have used his terms, if they did not start out with his 
presupposition. Valeri has made a mistake he himself warns against: 
"one unconsciously attributes to these writings our own principles of 
organization and criteria of intelligibility" (xxviii) . 

Similarly, in accordance with his separation of human beings from 
nature and with his philosophical orientation, Valeri seeks to establish a 
nonnatural or "supernatural," invisible, immaterial realm or dimen- 
sion, just like the one in Western thinking: for example, " 'supernatural' 
and 'natural' or rather invisible and visible" (92) . Valeri's arguments for 
this point are derived from his theory. For example, Hawaiian gods 
"retain a fundamental feature of the concept: nonempirical, transcen- 
dental reality. Thus, in principle, they cannot be confused with those 
among their instantiations . . . that are supposed to empirically mani- 
fest the god's properties" (32; see also 34, 261, 365 n. 23). In the next 
paragraph, he admits that this view cannot be found in Hawaiian texts, 
where they are always so "confused." 

Valeri's introductions of Western philosophical terms— introductions 
without argument — are used to support this view. Moreover, Valeri's 
language is very irregular when explicating it. For instance, he writes 
not only that mana, the gods, and "the divine" are "invisible" (89, 99, 
152), but that the alii are as well, giving such arguments as the fact that 
they were not usually seen by the commoners, that they went out at 
night, and so on (147, 150, 300-301): "It is only on this occasion that the 
commoners can see the most sacred ali'i . . . who are invisible through- 

Book Review Forum 129 

out the rest of the year" (380 n. 10). Similarly Kahiki is called "invisi- 
ble" apparently because it can't be seen from Hawai'i (8-9). 

Kahiki must be so treated — must be placed in a transcendental 
dimension rather than be accepted as a distant land within this uni- 
verse — because of the requisites of Valeri's theory: he uses this two-level 
or two-dimensional view of reality to interpret ritual. Ritual is "visible 
symbols" and "invisible realities" (131; cf. 56, 69, 215, 274, 300); "the 
gods' point of passage from transcendence to immanence" (250; also 
252). The offering is "the transfer into transcendence, outside the 
empirical world" (76). Such terms are used also in describing the ritual 
as thought process (323-325; also 268). 

Objections can be raised to the application of this view to Hawaiian 
culture. The Kumulipo pictures the universe as a single whole, not split 
into natural and supernatural dimensions. Gods do not have to come 
from another dimension to meet their devotees; they can come from the 
wao akua, "the uplands of the gods," which are found on each island, or 
from whichever part of the land or the sea they reside in. Moreover, I 
have seen no evidence that Hawaiians had the concept of immateriality. 
Gods, ghosts, and spirits — both in literature and contemporary testi- 
mony — have a body, which can be seen and even felt by those who are 
sufficiently talented or trained (for instance, Pukui and Elbert 1971, 
s.v. " 'ike papalua," "second sight"). Cupped hands are used to catch 
souls (Pukui, Haertig, and Lee 1972, 1:194). Such points could be 

Valeri's view introduces distortions at every level. For instance, he 
writes that at the hale mua the family "enters into relation with another 
aspect of the 'outside,' of the world that transcends the household: the 
gods" (174). In all my reading and listening, family gods appear very 
much part of the household, true family gods. 

Having established to his satisfaction a supernatural dimension, 
Valeri characterizes it by using concepts well known from the history of 
religion. He draws "some preliminary conclusions on the Hawaiian 
notion of akua, 'divine,' 'deity.' This notion is clearly characterized by 
two dualities. The divine manifests itself in both . . ." (31). The word 
akua can be used as a noun, "god," and as an adjective, "godly." But it is 
never, to my knowledge, used as an abstraction, "the divine." Valeri is 
introducing a non-Hawaiian idea, which he uses widely in his book, 
either in the vague sense of anything pertaining to the gods (e.g. , 31, 88, 
90, 153, 262) or as a generalized, undifferentiated "divine power": "a 
category of divine power exists that is more encompassing than the indi- 

130 Pacific Studies, Vol. 10, No. 2— March 1987 

vidual gods. From its point of view the individual gods are particular- 
izations associated with certain states of the process of transformation of 
the divine as it occurs at different levels of the ritual cycle" (288; see also 
78, 89, 215). As in Hinduism, the gods emerge from an undifferentiated 
divine and merge back into it. 

Valeri seeks to identify this concept with the Hawaiian po, "night," 
from the Kumulipo: "the divine coincided first with the undifferen- 
tiated principle Po. . . . This identification of the undifferentiated 
divine with Po . . ." (7); 39 "The closest approximation to a supreme 
divine principle found in Hawaii is Po, the undifferentiated creative 
origin of the cosmos, which continues to exist in transcendence as its 
perennial source" (35; also 215), an idea similar, for example, to the 
Thomist description of God as creator and sustainer of the universe. 

Valeri's Po, as the divine, differentiates itself both into the generators 
of the Kumulipo (5, "manifestations of the generative principle Po") 
and apparently into the major gods: by "producing the first man," the 
divine transforms itself in that "from now on it will be constituted by 
personal, anthropomorphic gods such as Kane and Kanaloa" (7). 40 

Valeri has two texts to support his view that Po is "the undifferen- 
tiated divine." The first text is the last line of the stanza used in the first 
four sections of the Kumulipo: " 'O ke akua ke komo, 'a'oe komo 
kanaka" ("It is the god who enters, the human being does not enter," my 
translation; 4, 7, 216, 222). Valeri bases his interpretation of this line on 
the idea that akua means "the divine," his abstraction: " 'the divine 
enters, man cannot enter'. . . . Being entirely divine, nature entirely 
excludes man" (7); "The Po period is thus entirely divine" (4). 

Valeri's argument depends, among other things, on whether the line 
from the stanza can be applied to Po. But when first used (line 39), that 
line is twenty-five lines away from the last mention of po (line 14). 41 
The immediate context of the line in question is the stanza itself, the 
first line of which gives a clear sexual meaning (Chariot 1977:499-500; 
1983a:49-52). The line could not, therefore, be applied to po without 
further argument. 

Valeri's only other text is the traditional expression "mai ka po mai" 
"from out of the night" (V: " 'out of the unseen' [out of the 'night']"), 
which, he states, "refers to anything of divine origin or 'supernatural' " 
(350 n. 3). Neither of his references supports his view or use of "super- 
natural." Moreover, Valeri's interpretation of the phrase would still not 
support the idea of Po as the "undifferentiated divine." 

Indeed, how could Po be undifferentiated if the god or even "the 
divine" "enters" into it? when earth, sky, moon, sun, slime, and so on, 

Book Review Forum 131 

have already been mentioned as existing in earlier lines? Far from refer- 
ring to a single, all-encompassing, undifferentiated principle, po is 
being constantly paired — with ao in the structure of the whole chant, 
with lipo in lines 7-8, and with la, "day," in line 10. This use is con- 
gruent with that in creation texts from the Society Islands and elsewhere 
(Chariot 1987b), which Valeri ignores since he makes almost no attempt 
to relate Hawaiian religion to its Polynesian background. 

Valeri has even less foundation for his view of Po as the "creative ori- 
gin of the cosmos" (35). The word po appears first in line 5 — after a 
description of the turning of the earth and sky and the sun being in 
shadow to illuminate the moon — and continues being mentioned along- 
side other elements (lines 6-10). Valeri himself sees those lines as identi- 
fying "the 'source' or 'origin' (kumu) of Po" (4) (an interpretation with 
which I disagree). The generation of Po is seen by Valeri as sexual: lines 
1-2 are "the Hawaiian equivalent of the marriage of heaven and earth"; 
walewale (line 6) refers to part of childbirth. Despite all this, Valeri 
then writes: "In turn Po engenders two forms that exhibit the first and 
fundamental biological difference, sex." In other words, despite all he 
has said, Valeri still uses Po as the origin and sees that origin as presex- 
ual; sex comes after Po. Indeed, he uses Po in this way throughout his 

Valeri is doing no less than replacing a two-source, sexual, genealogi- 
cal origin of the universe — the mating of the earth and sky — with a one- 
source, presexual one. His view can be arrived at only through his argu- 
mentation. He himself admits that "Po and its immediate specifications 
are not personified and do not receive a cult" (35; see also 215). The 
mating of earth and sky and the dualistic, sexual view of the universe 
are, however, richly in evidence: "Uwe ka lani, ola ka honua" ("The sky 
weeps, the earth lives," my translation; Pukui 1983, No. 2888; Chariot 
1983a, 49-52; 1983b). Sky and earth are proverbially considered the 
ultimate framework of the universe, for example "He lani i luna, he 
honua i lalo" ("A sky above, an earth below," my translation; Pukui 
1983, No. 718). Valeri himself describes "above and below, inland and 
sea" as "the two pairs of opposites that together encompass all that 
exists" (146). 

Because Valeri is replacing this two-source origin with a single-source 
one, he cannot use sex and procreation. He must use "creation" or "pro- 
duction": "the entire land, indeed all of nature used by man, is pro- 
duced by the gods and hence ultimately belongs to them" (156; see also, 
e.g., 7, 75). The Kumulipo is, however, a chant of the procreation, not 
the creation, of the universe. There is nothing other than late, biblically 

132 Pacific Studies, Vol. 10, No. 2— March 1987 

influenced Hawaiian texts to compare with the extended creationistic 
systems of Samoa and the Society Islands (Chariot 1987b). Hawaiians 
knew the god associated with those systems — Tangaloa/Ta'aroa/Kana- 
} oa — but, according to the evidence, never used creation by gods except 
in the limited sense of rearrangements of existing materials. 42 

In relating his undifferentiated single source to the gods, Valeri char- 
acteristically chooses the four masculine ones he regards as principal, as 
seen above. He divides the pantheon into male and female. In this pan- 
theon, KG "encompasses all the properties of the masculine gods," and 
Hina, "all the feminine attributes" (12; see also 13). Since goddesses will 
later be described as "marginal" to the pantheon, Valeri has managed to 
depict the pinnacle of Hawaiian religion as a masculine creator-god, a 
point, however, that he does not develop, as stated earlier. 

Valeri thus downgrades the role of the female in Hawaiian religion 
from being half of the pair that generates the universe. He consequently 
downgrades Hawaiian goddesses and women, imposing on them an old- 
fashioned Western image: "dancing is an activity in which women have 
a predominant role" (111); "that preeminently feminine function pro- 
creation, especially childbirth" (111); "women have a privileged rela- 
tionship with the female deities of sorcery . . . [as prophetesses and 
mediums]. Their mediating role thus takes on a typically feminine 
form: they are possessed, penetrated by the deity who speaks through 
them" (112); "superiority of action over passivity"; "Action is conceived 
as a masculine quality in contrast to feminine passivity. . . . Women are 
relegated to the unmarked, 'passive' category" (114); "the occupation of 
a young woman is to procreate, which in Hawaiian culture implies all 
that relates to seduction, in which it is said that women play a more 
active role than men. . . . This explains why properly feminine activi- 
ties are making ornaments . . . and clothing, chanting, dancing, and 
other activities that promote eroticism . . . compose and chant the 
mele inoa 'name chants,' with their deliberately erotic content and even 
the mele ma'i 'chant [praising] the genitals' " (123). This picture 
depends more on Western views of women than on evidence about 
Hawaiian ones. In fact, all the activities mentioned above, except child- 
bearing, were performed by men as well. 43 

In Valeri's picture, women constitute ritual impurity and pollution: 
"purity is an essentially masculine property, while impurity is essen- 
tially feminine" (112; also 18-19), thus the "global inferiority of women 
relative to men in the sacrificial system" (113). 44 This view influences 
Valeri's interpretation of Hawaiian ritual, 45 and he often appears to 
argue against his sources, imposing a one-source picture upon the con- 

Book Review Forum 133 

siderable evidence for a two-source, sexual ritual (e.g., 206, 217, 219- 
220, 282, 288; cf. 302-303). He dismisses without argument Kamakau's 
statement that the images on the left side of the temple fence and to the 
left of the altar were female, although this receives some support from 
Cook's journal (238, 245). When a source states that priestesses partici- 
pate in purification rites, Valeri responds, "Perhaps this rite reunites the 
sexes to represent their subsequent separation more emphatically" (389 
n. 53). This negative view of women enters even into speculation on 
details (e.g., 277). 

This imposition of a one-source view can be found in Valeri's inter- 
pretation of the Makahiki ceremony. Valeri correctly interprets the 
foundation myth, "the key to several aspects of the Makahiki ritual," as 
representing "the marriage of heaven and earth" (215). This is the basis 
for Valeri's holding "the identification of Lonomakua with a heavenly 
god uniting with the feminine earth" (214). Lonomakua is represented 
by a long pole with a crosspiece. Valeri writes, "Thus we ask ourselves as 
well if the long pole topped by a rounded head, which represents the 
god, is not a phallic symbol" (214). This pole is manipulated in various 
ways during a section of the ritual, held upright and laid down (208, 
222) . One would expect these points to be explicated in detail in Valeri's 
interpretation of the ritual (222-224) . However, there is no sexual refer- 
ence made to the pole and only passing reference to the sexual aspect of 
the ritual and its founding myth (222, par. 5; 224, par. 3). Valeri's char- 
acteristic emphasis is on the single male god Lonomakua, "manifesting 
that the earth and its products belong to the god that has produced 
them . . . the producer-god" (222); "Lonomakua is conceived of as the 
father or producer of cultivated plants," and so on (216). "Lono in the 
Makahiki rites" represents "all of the divine in a relatively concrete 
form" (215; see also 216) and is identified with "the divine origin" (216). 
A one-source picture is thus imposed on the evidence for a two-source 

In accordance with his one-source, male-god sacrificial theory, Valeri 
follows Frazer and Sahlins in interpreting the Makahiki ritual by the 
idea of the death and rebirth of a god (e.g., Sahlins 1985, 104-134; 
compare Daws 1968a, 26-27; Daws 1968b; Jean Chariot 1976, 81-96). 
This idea and its application as a model in various contexts are well 
known from the elaborated theologies and rituals of other cultures. It is 
surprising, therefore, how little evidence exists for its use in Hawai'i and 
in the rituals Valeri is interpreting. In the kali'i ceremony a spear is 
thrown at the king and misses. Another man touches the king with 
another spear. Valeri follows Frazer in interpreting this as "the king's 

134 Pacific Studies, Vol. 10, No. 2— March 1987 

'execution,' symbolized by the spear that touches him" (225; also 211- 
212, 285). No textual evidence supports this interpretation. The action 
could just as easily be viewed as a demonstration of the king's invulnera- 
bility, as a gesture of surrender, and so on. 

The above interpretation is, however, the point of departure for 
Valeri's application of the death-rebirth model to the god: "This rever- 
sal implies that Lonomakua is killed in turn, which is what happens 
immediately after the king's symbolic execution. In fact, conquered 
after a sham battle between the king's warriors and Lonomakua's 
defenders, the 'Makahiki gods' are brought into the king's temple and 
dismantled for storage" (226). 46 I cannot see how putting the images 
away after the ceremony is over implies that the god or gods are being 
killed. Again, no texts can be used to support this view. On the contrary, 
Lono is said to return to Kahiki (226). Moreover, the images of other 
gods are dismantled (392 n. 112), but Valeri says nothing about those 
gods being killed. Similarly, even the manipulation of the phallic pole is 
given a death-rebirth rather than a sexual interpretation: it is "placed in 
a horizontal position. Lono is thus 'beaten' and 'overthrown,' perhaps 
even symbolically killed" (222) . 

Valeri then argues for extending the death-rebirth idea to Ku and the 
temple ceremony (266, 285-288) — "the dubious world of eternal life 
that is in fact a world of sacrificial death" (326) — but his arguments are 
again tenuous. For instance, a hala lei is put around the neck of the god 
and the king. Hala can be used as a symbol or in wordplay for death. 
"Thus the king and god 'die' in one form in order to assume a superior 
one" (288). But in another historical account of the ceremony, a differ- 
ent lei is used. 47 

Again, one would expect textual evidence in this case because when 
Hawaiians want to say that a human being or a god is born or dies, they 
have no trouble doing so. 48 The problem to be solved is not, in my opin- 
ion, how to impose a death-rebirth theory where there is so little evi- 
dence for one, but, on the contrary, how to explain why the Hawaiians 
made so little of that ancient and widespread idea. There are indica- 
tions that they and other Polynesians were acquainted with that idea: 
for instance, the Hawaiian story of Ku becoming a breadfruit tree 
(Green and Pukui 1936:127), the Samoan story of Sina and the eel, and 
perhaps aspects of the practice of sacrifice. 49 That Hawaiians made such 
sparing use of that powerful image could be due to many factors: lack of 
winter and spring, planting obviously living taro-tops rather than dead- 
looking seeds, etc. I myself would see the reason in the strong and conse- 
quent dualism of Hawaiian thinking. Life and death are conceived as 

Book Review Forum 135 

real opposites. For instance, Hawaiians never made wordplays with 
two senses of make, "desire" and "death." When, at a conference I 
attended, a Western poet used the pun in a love song on a Hawaiian 
theme— "You're desire, you're death"— the Hawaiians listening were 
horrified. Hawaiians, as far as I can see, do not see death in life and life 
in death. They see life as health and vigor and joyous sexuality, and 
death as the opposite. Hawaiian ritual ideas and practices can be 
interpreted from that point of view, for which much evidence exists, 
and placed against an inherited background in which the death-rebirth 
idea is still perceptible. Hawaiian ideas and sensibilities should not be 
absorbed into foreign ones, but appreciated for their own special quali- 
ties and as the result of their own special development. 

By dint of much theorizing and extrapolation from few indications in 
the Hawaiian sources, Valeri has constructed a whole theology of sacri- 
fice and ritual: one-source, male-emphatic, creational, and spiritualis- 
tic, with divisions between human beings and nature and between the 
natural and the supernatural, and with major use of the death-rebirth 
model. The Hawaiian sources themselves abound in inescapable evi- 
dence of a very different theology: two-source, male-female, procrea- 
tional, physical, with an all-encompassing cosmos and a strong empha- 
sis on life over death. Valeri deals with this gap between his theory and 
the evidence just as he has handled earlier ones: he posits two different 
systems, a "genealogical system" connected to impure women and a 
"sacrificial system" connected to pure men (113); an "opposition 
between sexual reproduction, which is primarily associated with the 
feminine pole, and the sacrificial reproduction of social units with their 
natural correlatives, which is primarily associated with the masculine 
pole" (123-124; also 128). The former is characteristically subordinated 
to the latter; "Men's superiority to women, then, expresses only the 
superiority of a sacrificial relationship with the gods over a purely gene- 
alogical relationship with them. . . . sacrifice is superior to genealogy" 
(113-114). According to Valeri, female sexual reproduction is possible 
only because of male sacrifice: "the 'pure,' that is, nonsexual, reproduc- 
tion of the species as a concept is the sine qua non for the 'impure' (sex- 
ual) reproduction" (330); "fertility is actualized only by men . . . the 
ideal reproduction of the species by men translates into its empirical 
reproduction by women" (331). This male control can be seen in the 
king's regulation of genealogies (157-158). 50 After the abolition of the 
sacrificial system, "the hierarchy survived only in its genealogical form 
and consequently in its female-centered mode of reproduction. It is no 
accident, then, that female, not male, chiefs played the most important 

136 Pacific Studies, Vol. 10, No. 2— March 1987 

political roles after the abolition of the Old Regime" (128). Valeri pro- 
vides as a reference for this point the whole first volume of Kuykendall's 
The Hawaiian Kingdom. 

For anyone desiring to provide a coherent account of a logical system, 
such a division — along with the earlier ones noted — must pose a prob- 
lem. To solve it, Valeri elevates the word "creativity" to the role of a 
unifying term. In discussing the Makahiki festival "at a somewhat 
abstract level," Valeri discovers "the most general attribute of humans: 
creative activity as such" (233). This attribute "finds its objective correl- 
ative in the renewed creativity of nature during the season of the festi- 
val." Valeri is using the word "creativity" as a synonym for "procreativ- 
ity," an odd use influenced by the creationism discussed earlier. The 
word so used provides him with an anthropomorphic bridge between 
"man" and "nature." 

But Valeri goes even further, "Po and its immediate specifications 
. . . can be viewed as the projections onto the most general concept of 
nature of the most general aspect of the human species: pure activity, 
pure creativity. In the latter aspect the cosmos and the human species 
coincide and are therefore indistinguishable" (35). Students of the his- 
tory of religions will be interested to see Hawaiian religion interpreted 
as the Vedanta: Atman and Brahman, Soul and Cosmos, Subjective and 
Objective Reality are one. 51 

Valeri's Kingship and Sacrifice takes its place in a long line of works 
that have understood Hawaiian religion from foreign religions and 
theories, be they Chaldean, Egyptian, Christian, psychic, or psychiat- 
ric. Acknowledging in key places that his view is not supported by the 
evidence and that the weight of evidence supports in fact a different 
view, Valeri uses tendentious interpretations, omissions, and tenuous 
arguments to theorize his way to a counter-system of Hawaiian religion, 
one combined from well-known elements drawn from the history of 
religions. In so doing, he — like others before him— leaves out those ele- 
ments that are special and, I believe, valuable in Hawaiian religion: its 
strong sense of individuality and personality; its capacity for reverence 
and awe before the godly, human, animal, vegetable, and elemental; its 
sense of the interrelatedness of all things, including human beings; its 
placing of human beings within rather than above the universe; its 
understanding of everything in physical terms; the integrity of its search 
for wisdom through all the divisions created by Western culture; and so 
on. Bereft of such characteristic elements, Hawaiian religion ceases to 
be a challenge to Western thinking and becomes a mere example of reli- 
gious themes available elsewhere. 

Book Review Forum 137 

Hawaiian religion can be seen as itself only if looked at closely and 
carefully, that is, following scholarly rules of interpretation and argu- 
ment. Such rules must be formulated for each literature, and much dis- 
cussion is necessary, for instance, on uses of Polynesian genres and on 
the proper method for understanding Polynesian wordplay. Works of 
synthesis or those dealing with broad subjects should follow intensive 
and detailed philological studies. Valeri's book is valuable among other 
things for raising inescapably such questions of approach and method, 
an essential step in the development of the field of Pacific studies. 

Appended Note: Valeri's Criticisms of my 
"The Use of Akua for Living Chiefs" 

Valeri discusses the question of "The Divine King" (142-145) and offers 
arguments against an appendix of my book The Hawaiian Poetry of 
Religion and Politics, "The Use of Akua for Living Chiefs" (Chariot 
1985, 31-35), which he read some years ago in typescript. I was de- 
lighted to receive such an early response to my call for a general discus- 
sion of this question and take this opportunity to continue it. 

Unfortunately, Valeri begins by misstating my position: Chariot "has 
maintained that Hawaiian ali'i were not traditionally called akua" 
(144). As indicated in the title of my appendix, I am speaking of the 
application of that word only to living chiefs. After death, ali'i can 
become gods. This misunderstanding of my position misdirects Valeri's 
discussion. 52 

In my appendix I noted that the commonly accepted notion that 
high-ranking Hawaiian chiefs were called gods during their lifetimes 
rested on nineteenth-century prose accounts, such as those of Malo and 
Kamakau. Abundant evidence of this practice can be found during the 
period from the death of Kamehameha I through the early missionary 
period. Earlier evidence has not yet been found, which is surprising 
because there are many genres, such as laudatory chants, in which such 
an appellation should have been used and is in fact used in the post- 
Kamehameha period. Because nineteenth-century prose historical ac- 
counts could easily have been influenced by the post-Kamehameha 
practice, I looked for evidence in arguably earlier chants. I studied in 
some detail the relevant texts in the Chant for Kualii, which has been 
cited as an example of the practice in question. I found that the direct 
applications of akua to Kuali'i occurred in Kamakau's prose accom- 
panying the chant and that various problems attended the use of the 
word in the chant. I therefore drew the tentative conclusion that the 

138 Pacific Studies, Vol. 10, No. 2— March 1987 

application of akua to living chiefs as found in the post-Kamehameha 
period and reflected in the nineteenth-century prose accounts was 
either a late innovation or a late extension of a genuine but restricted 
early practice. I cited several texts that suggested that Kamehameha I 
might have been the source of such an extension or innovation, as he 
was of so many others. This would have been one of the several religious 
changes accomplished during his lifetime and after his death (Chariot 
1983a, 26-29, 147-148; Chariot 1985, e.g., 5-6, 34, 55). Although 
Valeri indulges in name-calling on this point, he himself refers on sev- 
eral occasions to Kamehameha's important innovations. 53 

In arguing for his own position, 54 Valeri has recourse to nineteenth- 
century prose accounts (several of which were cited previously in my 
appendix), ignoring the general objections I raised against them. He 
also ignores points I made against specific texts when he seeks to use 
them for his own position. For instance, Kamakau writes that some 
people worshipped a victorious chief me he akua, "as a god" (Chariot 
1985, 30; Valeri 1985, 143 "like a god"). Valeri is clearly not given pause 
by me he, "as" or "like," although a text of Kepelino's, which he cites 
(131), has a parallel use that shows its significance: the diviner was "me 
he mea atua la, 'like a god' " (Kepelino 1977, 60-61). Neither Valeri nor 
anyone else argues that diviners were called gods. 

Similarly, in using a chant text cited by me (Chariot 1985, 31; 
Fornander 1919-1920, 6:387-388), Valeri refers to line 300, which 
mentions Kamehameha's waiakua, "godly blood" (143; V. "divine 
blood"), but leaves out the previous lines, which mention his waikana- 
ka, "human blood" (line 299), and the fact that he is indeed he kanaka, 
"a human being" (line 297). 

Valeri offers two arguments that do not require a long refutation. 
Texts about chiefs being descended from the gods do not prove that they 
were called gods during their lifetimes, nor does the fact that they were 
given the names of gods. After all, Hispanics can call sons Jesus. 

The decisive evidence on which to judge the difference of opinion 
between us can be reduced to a few texts from Hawaiian chants: 

1. In his remarks on the Chant of KualVi (143, 145, 392 n. 98), Valeri 
does not take account of my objections to its use for his purpose (Chariot 

2. Valeri cites a chant in which he claims Kakuhihewa is called " 'he 
akua 'olelo,' 'a god of speech.' " The chant is in Emerson's notes to Malo, 
which Valeri describes as "a mixture of data of great value and 
unfounded or misunderstood information" (xxiv). Emerson himself is 
vague about the provenance and reference of the chant (Malo 1951, 

Book Review Forum 139 

200). It contains in fact the word mol for "king," which Valeri follows 
Stokes in characterizing as a usage introduced after 1842 under foreign 
influence (370 n. 36; also, 397 n. 192). The phrase under discussion 
refers not to Kakuhihewa but to the chief mentioned in the previous 
line, "Ka-ua-kahi-a-ka-ola," who seems to be a child of Kakuhihewa by 
"Kanui-a-panee." 55 In any case, the fact that the word 'olelo qualifies 
akua — as well as the uncertainty of any identification, reference, and 
interpretation of the chant— infirms this text for Valeri's purpose of 
demonstrating an absolute, not a qualified, application of the word 
akua to a living chief. 

3. Valeri claims that, in Ke'aulumoku's Haul Ka Lani (Fornander 
1919-1920, 6:408, line 734), "the ali'i of the district of Hilo are ironi- 
cally referred to as akua" (143). The line reads, however, "Liu na maka 

na akua i ka paakai, 'Blinded are the eyes of the gods with salt.' " No 
mention is made of aWi. Valeri is taking his interpretation without sup- 
porting argument from note 1 to the Fornander text. The line makes 
perfect sense when taken literally. 

4. The saying, He akua na alii o Kona, "The chiefs of Kona are 
akua" was mentioned by me with two parallels (Chariot 1985, 31-32). 

1 followed Dickey, who translates akua as "ghosts." The first parallel I 
cited has not been translated. The second (Kahiolo 1978, 61) also trans- 
lates the term as "ghosts." Valeri polemicizes against the translators of 
Kahiolo (370 n. 37) for "following Dickey, without, however, naming 
him." The translators were in fact working from the best-known refer- 
ence, the parallel in the Fornander version of the saga of Kamapua'a 
(Elbert 1959, 222-223), which I unfortunately did not cite. The transla- 
tors of Kahiolo very correctly describe their relation to the Fornander 
translation in their preface (Kahiolo 1978, ix-x). The above consensus in 
the translation of akua as "ghosts" in this line is based on its context in a 
taunting chant (or teasing in Dickey): Kamapua'a is taunting Pele by, 
among other things, using a typical description of ghosts against her 
(compare Fornander 1919-1920, 6:370-371). Such a chant is not the 
proper context for a glorifying reference to chiefs as gods, but for a 
taunting one, applying to the chiefs the Hawaiian idea of miserable and 
helpless ghosts feeding on scraps. 

Not one of the texts produced up to now in this discussion has shown 
convincingly that living chiefs were called akua in pre-Kamehameha 
times. But even if all the texts Valeri cites were accepted, they would 
still be surprisingly few for a point so important and useful in praise. 
One cannot argue with Valeri for earlier times that "it can be safely 
assumed that this usage really existed because it follows necessarily from 

140 Pacific Studies, Vol. 10, No. 2— March 1987 

the attribution to ali'i of the fundamental properties of the divine" 
(145), the prelude to a discussion I have criticized at some length above. 
Theories must be based on evidence, not the reverse. Until better evi- 
dence is found, scholars run the risk of projecting a later practice onto 
earlier times in assuming that the nineteenth-century accounts of this 
particular practice are accurate for earlier periods. 


1. All references are to Valeri 1985, unless otherwise noted. Valeri has treated certain 
subjects and themes found in this book somewhat differently in the articles listed in his 
bibliography. I have not used them for this article, which discusses his most recent work in 
publication, if not in composition. 

I read a preliminary version of this paper to the Humanities Forum, Institute of Culture 
and Communication, East-West Center, chaired by Wimal Dissanayake, and received 
valuable comments and criticisms. I thank also Jocelyn Linnekin and Jean-Paul Latouche 
for their detailed criticisms of the whole typescript. 

2. Ortiz 1969, see Index, s.v. "Levi-Strauss, Claude." Valeri states that "the disjunction 
between the raw offering and the cooked offering" cannot be found in Hawai'i (123). 

3. An example of such practice in regard to a female ali'i can be found in a text cited by 
Valeri for another reason, but he ignores it as evidence against his point (166; Fornander 
1916-1917, 4:540-541, 544-545). In a later passage (150, end of par. 3), Valeri offers fur- 
ther references, which arguably support the virginity of the male, but there is again no ref- 
erence to purity. See also Elbert 1956-1957, 69:342. 

4. The words from the English translation "free from the taint of kauwa blood" and 
"regarded as a defilement" (Malo 1951, 71) do not appear in the Hawaiian text (Malo 
n.d., 83-84). The first phrase has simply been added. The second is a misreading of "ua 
kapa ia ka poe kauwa he palani, he hohono ke ano" ("Kauwa were called palani fish, a 
type that has a bad odor," my translation). Genealogists try to hoomaemae, "cleanse," 
chiefly lines of such connections. The word ho'oma'ema'e can be used of ritual cleansing of 
pollution, but, not being a technical term, need not have such a sense in this context. It is 
the only word in the text that could offer support to Valeri's position. 

5. Confusingly, Valeri later speaks differently of immobility (301-302): the wild chiefly 
woman is "immobilized" by being wrapped, which, Valeri states, makes her productive; 
but for that — as seen in his immediately previous paragraphs — she must surely be 

6. Compare Fornander 1969, 1:92-93, quoted by Valeri (200). 

7. Valeri thus translates the name of the god on p. 175. Valeri's view (expressed on p. 6) 
would make it very hard to translate lines 1 1 1-1 12 of the Kumulipo. 

8. See also pp. 86, 93 par. 6, 390 n. 80, 391 n. 86; on kahea, see 379 n. 3, 398 n. 201. 

9. For other examples of such arguments, see pp. 279 par. 2, 391 nn. 86, 96. 

Book Review Forum 141 

10. For other problematical arguments of various kinds, see, r.g., pp. 60, 99, 232, 270, 
887 par. 4, 302-303, 322, 323 par. 1, 324-325, 366 n. 26, 371 n. 49, 373 n. 69, 393 n. 124. ' 

11. See also the hypothesis about mana (99) and how it is used (101, par. 7). Also pp. 251 
par. 1, 252 par. 2, 270 par. 2. 

12. "It seems that the opposition akua/kanaka is a relative one and that certain men may 
be called the gods of others. . . . Probably because he does not take this relativity into 
account, John Chariot ... has recently maintained . . ." (144). I do not in fact agree 
with Valeri's view of the distinction. See Appended Note above, p. 137. 

13. Similarly on p. 252, the points made are to be confirmed by the analysis of the temple 
ritual I discuss below. 

14. See pp. 32, 34 par. 6, 215 par. 6, 225 par. 4. Compare p. 388 n. 41. This is true for a 
number of small points as well (e.g., 166). 

15. See also pp. 258-262, especially 260 n. 61, 326. 

16. See also pp. 304-305, 346; cf. 42, par. 3. This view also influences Valeri's interpreta- 
tion. For instance, he speaks of "collective recognition" of the success of the ceremony; "It 
is all of society that decides" (304; also, 305, 394 n. 140), as seen in the fact that the crowd 
must keep absolute silence during the ceremony and could render it invalid by making or 
reporting a noise. It should be noted, however, that anyone who made a noise was killed. 

17. Valeri's sensibilities seem Western: e.g., on white as the color of purity (52, 86); on 
sweet potatoes and excrement (123); on women composing "even"' genital chants (123). 

18. "Consanguinity implies identity" (163; see also 93). Each chief was proudly conscious 
of his many differentiating names, kapus, traditions, and so on, which were celebrated in 
story and song. 

19. Valeri himself seems to say so (168, par. 5). His previous reference (166) does not con- 
tain the words he underlines or the idea they express. 

20. A few lines down Valeri refers to the chiefs' "perfect self-control" (147); "the ali'i is 
divine as long as he acts like a god, as long as he manifests his perfection by not desiring or 
needing other human beings" (149); see also 166. 

21. See also 123; above, p. 110; Chariot 1985, 3 and n. 11, p. 10 and n. 57; Elbert 1956- 
1957, 69:341-345. Valeri writes similarly of prophets (139), but the most famous one, 
Lanikaula, was married and had children. 

22. "A deity is nothing but a reified representation of certain human properties. It follows 
from this that . . ." (46); "This is only logical . . ." (385 n. 5). 

23. See pp. 191-194, 203, 232, 254-255, 260, 280, 304, 322, 401 n. 239. 

24. E.g., pp. 25-27, 29-30, 112, 115, 119 and n. 6, 194-198, 225, 235-236, 248, 254, 351 
n. 18, 364 n. 15, 365 n. 23, 369 n. 24, 372 n. 62, 377 n. 24, 382 n. 32, 390 n. 77, 398 n. 

25. "These 'residual' persons are destructive as is any residue with respect to the system 
that produces it" (370 n. 31). 

26. The latter example is particularly instructive because the religious elements involved 
are similar. Compare, e.g., pp. 138-139 to Hoffman 1891, 156-162. 

142 Pacific Studies, Vol. 10, No. 2— March 1987 

27. His example is faulty: the chant lists various groups of gods named by number, and 

E ke kokoo-ha o ke 'kua 
E ke koo-lima o ke 'kua 

Oh association of four of the god(s) 
Oh association of five of the god(s) 

The chant is not referring to a single, overall supreme group, but to a number of groups. I 
do seem to remember the word being used as Valeri states, but have not found the text. For 
Trinitarian notions, see, e.g., Kepelino, in Beckwith 1932, 8-11, 14-15, 174-175. For such 
biblically influenced theologies, see Barrere 1969. 

28. See also pp. 19, 119. Valeri admits some examples are "due to Christian influence," but 
does not offer references for those he feels are not. He also recognizes other examples of 
biblical influence on Kamakau and others (353 n. 9, 358 n. 60, 377 n. 23). 

At one point, Valeri goes even further. Ignoring evidence for the supremacy of Kane in 
some areas (as seen above), he exalts Ku as the highest of the male gods, who "encompasses 
all the properties of the masculine gods" (12). Given the "marginal" position he accords to 
women in the "pantheon," this approaches a henotheistic position (12-13). Valeri does not, 
however, develop this point. 

29. Earlier, he claims this only of the "majority" (13). 

30. He connects Pele elsewhere also with vulcanism and dance (8). 

31. The Kamakau text cited, however, discusses only the statues of those gods and does 
not, therefore, support Valeri's point. 

32. Valeri seems to contradict himself, when he writes "as all other deities, the 'aumakua 
[sic] can appear in human form" (21), and then mentions "the links connecting various 
individual 'aumakua [sic] to a single natural species" (30). 

33. A useful list of categories is given in Fornander 1919-1920, 6:52-55, a text Valeri refers 
to (266). Valeri (10 and n. 6) ignores the gods who emerge with animal species in the 
Kumulipo before the birth of anthropomorphic gods and human beings (e.g., Klwa'a, line 
366). For gods with only animal bodies reported, see, e.g., Fornander 1918-1919, 5:366 
(shark); Green 1923, 16-17 (bird), 44-45 (caterpillars), 46-47 (squid); idem. 1926, 66-69 
(rat and owl); Green and Pukui 1936, 174-175 (squid), 176-177 (fish). Many contempo- 
rary Hawaiian religious experiences involve animals with no reported human form (e.g., 
Chariot 1983a, 22). Elemental gods include rocks and waves, Fornander 1918-1919, 
5:522-555 (waves). Sources differ on whether some gods possessed a human body as well 
as an animal body, for instance, the dog Pae or Pa'e (Green 1923, 48-49; cf. Green and 
Pukui 1936, 178). It would be systematizing to argue that if a human body can be discov- 
ered in one source, all others must implicitly agree with it. I sketch my own view of the 
subject in Chariot 1983a, 21-22, 146-147. I believe that anthropomorphism is a later ele- 
ment in Hawaiian religion that was applied secondarily to the older theriomorphic gods, 
partly through the identification of earlier animal gods with anthropomorphic ones; for 
instance, Tlioloa, "Long Dog," can become either Ku'llioloa or Kane'llio. 

34. Kahiolo 1978, 52-59, and the parallel passages in the two other major Kamapua'a 
complexes. In Chariot 1987a, I demonstrate that the oldest stories of Kamapua'a depict 
him as a pig and that his human body is a later development of the literature. 

Book Review Forum 143 

35. E.g., Green 1923, 43 (caterpillar); Green and Pukui 1936, 170-173 (eel and sea 
cucumber); Valeri 1985, 331 (mo'o). 

36. Similarly, Valeri rejects without argument a statement by Malo that nonanthropo- 
morphic statues were made, stating, "The fact is that all surviving images are anthropo- 
morphic" (9). In fact, a number of nonanthropomorphic, undeterminable, and unshaped 
stone gods can be seen at the Bishop Museum. Valeri places great theoretical emphasis on 
the use of such statues in ritual (72-73), going so far as to say "ritual is fully efficacious 
only when the god is present in an anthropomorphic, controllable form" (102). That was, 
however, not always the case, as he admits (103). For instance, the goddess Laka was 
represented on the altar by an uncarved block of wood. 

37. See also pp. 75-81; cf. 24: "the social universe encompasses the natural universe"; and 
1 19. On the general point of separation, see pp. 9, 18, 31 , 34, 48, 76. 

38. E.g., "By deanimalizing the animal that is the object of his fear, man deanimalizes 
himself (24). This idea is applied to ritual following Hegel (48). See also pp. 75-81. 

39. He speaks of Po as one of the "metaphors designating the divine origins" (8). Valeri's 
descriptions of Po often recall those of mana in the writings of other anthropologists. For 
my own view, see Chariot 1977, 498-500; Chariot 1983a, 124-125. 

40. Similarly Ku and Lono can "represent" "the divine" (215-216). For the connection 
between the divine and Po on this point, see p. 383 n. 48. Valeri's understanding of the 
chants cited in note 48 is again based on his theory. 

41. The po in line 37 of the Beckwith text reads pou in the manuscript (Beckwith 1972, 

42. The use of the word "creation" to designate all views or models of cosmic origins is an 
ethnocentrism widely and unconsciously perpetrated by Western scholars (e.g., Edmond- 
son 1971, 7, line 63; 11, lines 173-174). Valeri's text is a good example of the power of dis- 
tortion of such a use. Valeri's emphasis on creation — rather than procreation — entails his 
elevation of "sight and intelligence" to "what is most human" (252 and n. 44; also 324- 
325; but see 396 n. 174). He thus follows a characteristically Western line of argument. 
Significantly, Hawaiians had several differing terms for thinking. "Traditionally, the intel- 
lect and emotions were thought to exist in the intestinal regions" (Pukui, Haertig, and Lee 
1972, 1:155). See also Pukui and Elbert 1971, s.v. "na'au," "na'auao," " 'imi naauao"; 
Chariot 1983a, 29. Moreover, a major Hawaiian literary form was the mele ma'i, "genital 
chant," which celebrated the recipient's sexual and fertilizing powers, a logical genre for a 
culture based on genealogical thinking. 

43. Contrast the view of the "high position of women" in Elbert 1956-1957, 70:320; also, 

44. Valeri bases his ritual hierarchy on the idea of purity, but his discussion of purity— 
which he connects with completeness or wholeness of instantiation of type (e.g., 84, 88, 
92, 148-149, 271, 276)— is marked by inconsistencies. For instance, he writes of the king, 
"Most of the taboos surrounding his sacred person are intended to maintain this purity" 
(148); but later, "any taboo surrounding a pure sacred being has as its aim to protect his 
purity" (374 n. 83). 

Valeri writes (130-131), "What characterizes the sacrificer ... is that he is per- 
manently in contact with the god." (In fact, none of Valeri's references support that view. 

144 Pacific Studies, Vol. 10, No. 2— March 1987 

To "have" a god or to be "connected" with a god does not mean that one is "permanently 
in contact" with it. On the contrary, in prayers one must call on the gods to come.) Valeri 
goes on to say that the sacrificer participates "in this way in the god's nature and mana, 
even being his manifestation in a human form" (130-131). Yet, as Valeri shows, it is neces- 
sary for these "pure people (ali'i, kahuna)" before the temple ritual "to be purified anew" 
(259; also 256-258, 267). How could they be in permanent contact with their gods if they 
were in a state of impurity, even of relative impurity? 

Valeri does in fact speak at times in relative terms: "both what is marked kapu and what 
is marked noa can be pure or impure and are so only relatively" (259); "Kapu and noa are 
purely relative notions" (90; also 326, 330). But he can speak also in absolute terms: "at 
least some part of the divine must always remain kapu for the human sphere, otherwise 
the whole system would lose its fixed foundation" (91); " 'to pollute' would be a more 
appropriate rendering, since the land is made accessible to human use (noa) by being de- 
sacralized and losing its divine purity" (259). 

Despite his expressed separation of kapu-noa from pure-impure, Valeri can write as if 
pollution were the way to lift a kapu (e.g., 91-92, 259, 326). He does use the word "free" 
to translate noa when describing the rendering noa of a temple (327-329, 401 n. 247), not 
wanting apparently to suggest that the temple was polluted. But he does not hesitate to 
refer to the pollution of the earth in order to render it workable, as seen above (also 19, 
120-121). This does not accord with Hawaiian literary expressions of love for the land and 
of feelings of awe before its numinous quality, or with the considerable evidence for 
Hawaiian practices in regard to the land (e.g., 154, 348, 360 n. 75). 

The above Hawaiian view of the land is based on the genealogical, two-source, earth- 
sky picture of the universe. Valeri sketches an alternative view, "divine nature," based on 
his theory of a "producer god" (7, 75, 156). 

45. E.g., p. 326. This view influences also his interpretation of texts, as seen above, p. 
131. Also, in his interpretation of a farmer's chant, he misses the fact that one half of the 
god list is female (see above, p. 111). 

46. Also "the Makahiki gods are taken down, wrapped up, and stored in the luakini' 


47. Cf. p. 308. Valeri uses equally poor arguments elsewhere (278): 'Umi undergoes "a 
symbolic death (the death sentence that is not enacted)." Also pp. 232, 287. 

48. E.g., Kumulipo, lines 612-615; Valeri 1985, 310. There are stories of killing gods and 
ghosts by trickery. A human being can die, make, and then his ghost can be killed or eaten 
by another spirit, in which case he is make loa. 

49. See p. 359 n. 68 for other possible examples. 

50. The practice Valeri refers to was, however, limited in time and place. Valeri here as 
elsewhere exaggerates the control of the "king." For instance he states that noble rank 
depended on relation to the king's genealogy (157-158 and n. 69, 296-297). To do this he 
must admit to going beyond the evidence in Malo. Such a relation to the king can be found 
after Kamehameha I (Chariot 1985, 5-6 and n. 30), but arguments would need to be 
offered to revise the general view that individual chiefly families could demonstrate their 
rank from their own historical backgrounds, traditions, and genealogical lines. To give just 
one aspect of this, families could have prestigious kapu and kandwai that belonged to the 
family and were not connected to the "royal" line. By his use of the term "king" for the 
time before Kamehameha I, Valeri introduces a number of anachronistic elements. 

Book Review Forum 145 

51. The influence of Hinduism on Valerfs thinking has been noted earlier. Hinduism has 
also influenced the scholars who have influenced Valeri (64). Valeri does note differences 
between Brahmanism and Hawaiian religion (e.g., 66, 90, 92). 

52. See p. 145 par. 2; also 370 n. 37, where he seems to forget that "ghost" is a perfectly 
legitimate dictionary meaning of akua. 

53. "Chariot ends up treating Kamehameha as a culture hero who possesses quasi-super- 
natural powers of transformation and innovation" (145). This position is foreign to my 
thinking. On changes initiated by Kamehameha I, see Valeri, pp. 184, 222, 230-231, 369 
n. 21, 372 n. 62, 385 n. 73. 

54. Valeri is in fact inconsistent in his statements on the point. He states flatly, "the ali'i 
themselves can be considered deities" (44); but then he writes of "the king's 'divinity' (that 
is, of the fact that he is closer to the divine than any other human, and therefore akua rela- 
tive to others)" (142). Valeri wants to turn akua into a relative term, just as he has done 
with purity-impurity and kapu-noa: "the king is a manifestation of his gods and is there- 
fore himself a god relative to all other men" (145). 

55. In other sources, the nearest son's name I have found is "Kauakahinui-a-Kakuhi- 
hewa," whose mother, however, bears another name than that given in the chant (Fornan- 
der 1969, 2:274, 276) The nearest daughter's name given is that of the descendant 
"Kauakahikuaanaauakane" (Fornander 1969, 2:276; Kamakau 1961, 62, 74). A further 
descendant "Kauakahi-a-Kahoowaha" is claimed as mo'i, "king," of O'ahu and was father 
of Kuali'i (Fornander 1969, 2:277-278). The nearest name I have found is that of "Ka-ua- 
kahi-a-kaha-ola" (Sterling 1974, 37), a counselor from Kaua'i at the court of Kalani'opu'u 
and later Kamehameha. The description of a famous counselor and educator as a "god of 
speech" would be hyperbolic but appropriate. But the other information in the chant does 
not seem to fit. 


Barrere, Dorothy B. 

1969 The Kumuhonua Legends, A Study of Late 19th Century Hawaiian Stories of 
Creation and Origins. Pacific Anthropological Records No. 3. Honolulu: B. P. 
Bishop Museum. 

Beckwith, Martha Warrren 

1932 Kepelinos Traditions of Hawaii. Bernice P. Bishop Museum Bulletin 95. Hono- 
lulu: B. P. Bishop Museum. 

1972 The Kumulipo, A Hawaiian Creation Chant. Honolulu: University of Hawaii 

Chariot, Jean 

1976 Two Hawaiian Plays, Hawaiian English. Honolulu: Jean Chariot. 

Chariot, John 

1977 "The Application of Form and Redaction Criticism to Hawaiian Literature." 
Journal of the Polynesian Society, 86, no. 4 (December): 479-501. 

1979 "The Hula in Hawaiian Life and Thought." Honolulu Magazine, vol. 14, no. 5 
(November): 49-56. 

146 Pacific Studies, Vol. 10, No. 2— March 1987 

1983a Chanting the Universe, Hawaiian Religious Culture. Honolulu and Hong Kong: 

Emphasis International. 
1983b "A Pattern in Three Hawaiian Chants." Journal of American Folklore, vol. 96, 

no. 379 (Jan.-Mar. 1983): 64-68. 
1985 The Hawaiian Poetry of Religion and Politics, Some Religio-Political Concepts in 

Post-Contact Literature. Monograph Series, no. 5. Laie, Hawaii: The Institute 

for Polynesian Studies. 
1987a The Kamapuaa Literature, The Classical Traditions of the Hawaiian Pig God as 

a Body of Literature. Monograph Series, no. 6. La'ie, Hawaii: The Institute for 

Polynesian Studies, forthcoming. 
1987b "Four Society Islands Creation Texts." Journal de la Societe des Oceanistes, in 


Daws, Gavan 

1968a Shoal of Time: A History of the Hawaiian Islands. New York: Macmillan. 
1968b "Kealekekua Bay Revisited: A Note on the Death of Captain Cook." Journal of 
Pacific History 3:21-23. 

[Dibble, Sheldon, ed.] 

1838 Ka Mooolelo Hawaii. Lahainaluna, Hawaii: Mea Pai Palapala No Ke Kulanui. 

Edmonson, Munro S. 

1971 The Book of Counsel: The Popol Vuh of the Quiche Maya of Guatemala. Middle 
American Research Institute, Publication 35. New Orleans: Tulane University. 

Elbert, Samuel H. 

1956- "The Chief in Hawaiian Mythology-." Journal of American Folklore 69:99-113, 
1957 341-355; 70:264-276, 306-322. 

1959 Selections from Fornander s Hawaiian Antiquities and Folk-Lore. Honolulu: 
University of Hawaii Press. 

Fornander, Abraham 

1916- Fornander Collection of Hawaiian Antiquities and Folk-Lore. Memoirs of the 
1920 Bernice Pauahi Bishop Museum Vols. 4-6. Honolulu: B. P. Bishop Museum. 
1969 An Account of the Polynesian Race, Its Origin and Migratiotis and the Ancient 

History of the Hawaiian People to the Times of Kamehameha I. 3 vols. Rutland 

and Tokyo: Charles E. Tuttle. 

Green, Laura S. 

1923 Hawaiian Stories and Wise Sayings. Poughkeepsie, N. Y. : Vassar College. 
1926 Folk-tales from Hawaii. Poughkeepsie, N.Y.: Vassar College. 

Green, Laura S., and Mary Kawena Pukui 

1936. The Legend of Kawelo and other Hawaiian Folk Tales. Honolulu: Territory of 

Hoffman, W. J. 

1891 The Midewiwin or "Grand Medicine Society" of the Ojibwa. Seventh Annual 
Report of the Bureau of Ethnology to the Secretary of the Stnithsonian Institution 
1885-'86, pp. 143-300. Washington: Government Printing House. 

Kahiolo, G. W. 

1978 He Moolelo No Kamapuaa, The Story of Kamapuaa. Translated by Esther T. 
Mookini, Erin C. Neizmen, and David Tom. Honolulu: Hawaiian Studies Pro- 
gram, University of Hawaii. 

Book Review Forum 147 

Kamakau, Samuel M. 

1961 Ruling Chiefs of Hawaii. Honolulu: The Kamehameha Schools. 


1977 "Kepelinos 'Hawaiian Collection': His Hooiliili Havaii,' Pepa I, 1858." Trans- 
lated by Bacil F. Kirtley and Esther T. Mookini. Hawaiian Journal of History 

Kirch, Patrick Vinton 

1984 The Evolution of the Polynesian Chief doms. New Studies in Archaeology. Cam- 
bridge: Cambridge University Press. 

1985 Feathered Gods and Fishhooks: An Introduction to Hawaiian Archaeology and 
Prehistory. Honolulu: University of Hawaii Press. 

Linnekin, Jocelyn 

1985 Children of the Land: Exchange and Status in a Hawaiian Community. New 
Brunswick: Rutgers University Press. 

Malo, David 

n.d. Ka Moolelo Hawaii. Hawai'i-Pacific Collection, Hamilton Library, University of 

Hawaii. Photocopy. 
1951 Hawaiian Antiquities (Moolelo Hawaii). Translated by Nathaniel B. Emerson. 
Honolulu: B. P. Bishop Museum. 

Ortiz, Alfonso 

1969 The Tewa World: Space, Time, Being, and Becoming in a Pueblo Society. Chi- 
cago and London: University of Chicago Press. 

Pukui, Mary Kawena 

1983 'Olelo No'eau, Hawaiian Proverbs 6- Poetical Sayings. Bernice P. Bishop Museum 
Special Publication 71. Honolulu: Bishop Museum Press. 

Pukui, Mary Kawena, and Samuel H. Elbert 

1971 Hawaiian Dictionary. Honolulu: University of Hawaii Press. 

Pukui, Mary Kawena, E. W. Haertig, and Catherine A. Lee 

1972 Nana i ke Kumu (Look to the Source), Vols. 1-2. Honolulu: Hui Hanai. 

Pukui, Mary Kawena, and Alfons L. Korn 

1973 The Echo of Our Song: Chants 6- Poems of the Hawaiians. Honolulu: University 
Press of Hawaii. 

Remy, Jules 

1862 Ka Mooolelo Hawaii, Histoire de VArchipel Havaiien (lies Sandwich). Paris and 
Leipzig: A. Franck. 

Sahlins, Marshall 

1985 Islands of History. Chicago and London: University of Chicago Press. 

Sterling, Elspeth P. 

1974 Index to Ruling Chiefs of Hawaii by S. M. Kamakau. Honolulu: B. P. Bishop 

Sterling, Elspeth P., and Catherine C. Summers 

1978 Sites of Oahu. Honolulu: B. P. Bishop Museum. 

148 Pacific Studies, Vol. 10, No. 2— March 1987 

Response: Valerio Valeri 
University of Chicago 

John Chariot purports to have written a "critical review" of my book 
Kingship and Sacrifice: Ritual and Society in Ancient Hawaii. Critical 
it is, but a review of my book it is not. 

It is not a review of my book because Chariot does not do what a 
reviewer should do: summarize the whole argument of a book before 
evaluating it. It is also not a review of my book because, by sheer power 
of misunderstanding, gross manipulation, and outright misquoting, 
Chariot has managed to create a travesty that has little recognizable 
relation to my book. He has thus accomplished the rare feat of seeing 
neither the forest nor the trees. Since he is mostly concerned with the 
trees, my rejoinder will have to follow him on his elective ground; how- 
ever, I will repeatedly have to refer to the forest in order to indicate the 
true location of certain trees, or sometimes blades of grass that he mis- 
takes for trees. 

Before addressing Chariot's arguments I cannot avoid mentioning a 
rather unpleasant fact. Chariot seems to have been much disturbed by 
my criticism of one of his pet theories. As I shall show later at the end of 
this rejoinder, he has given me no reason to retract my criticism, but I 
find his accusation of having indulged in "name-calling" in the course of 
my criticism (Chariot 138) quite unacceptable. Far from being rude to 
him, in the acknowledgments I say, "I warmly thank him for his help" 
(Valeri xv). Since Chariot seems unable to recognize polite language, I 
have decided to be much less careful with my words in this rejoinder 
and to attempt to match, as far as I am able, the unpleasant tone of his 
prose. I may be allowed to observe that this false accusation is only the 
most offensive example of his systematic distortion or even falsification 
of my statements throughout his piece. 

Chariot begins by saying that there are many "inaccurate references" 
in my book and gives a number of examples that, presumably, he finds 
particularly blatant. Let me examine each of these examples in the 
order in which they are discussed by Chariot. 

1. My statement (V. 149) that "divine ali'i ... are obliged— men 
and women— to remain virgin until marriage" is not, says Chariot, sup- 
ported by the source that I quote at that page. He claims this because 
the source in question states that "girls were required to be virgin until 
the first planned union to conceive a child" and does not mention that 
the same rule was valid for men. However, Chariot must admit (charac- 
teristically in a footnote— C. 140 n. 3— which is then not taken into 


Book Review Forum 149 

consideration in his negative evaluation of my argument) that other 
sources that I quote on the following page (V. 150) "arguably support 
the virginity of the male" (C. 140 n. 3). 1 In other words, my statement is 
correct and my only fault consists in not having quoted all the evidence 
on page 149. Chariot also claims that the general statement that opens 
my discussion of this and related practices, namely that they are meant 
to preserve the purity of the ali'i, is not supported by the source that I 
quote on page 149 (a text from Pukui, Haertig, and Lee, whose volumes 
I will henceforth quote as NK 1 or NK 2) nor by any other source. In his 
opinion, "virginity was maintained ... for purely practical genealogi- 
cal reasons, not for the maintenance of ritual purity" (C. 108). There 
are two points here: one is my alleged use of Pukui, Haertig, and Lee to 
support the thesis of a connection between virginity and ritual purity; 
the other is the validity of this thesis as such. 

Concerning the first point, had Chariot quoted me in full, it would 
have been evident that my reference to Pukui, Haertig, and Lee was not 
meant to support my interpretation of virginity as a sign of ritual purity, 
but to document two specific facts: the taboo on intercourse between 
sacred ali'i and women of lower rank and the chiefly taboo on having 
sexual relations before the first marriage. Indeed, I write (V. 149): "The 
purity of sacred ali'i is preserved not only by the behavior of their infe- 
riors or rivals, but also by their own comportment. For example, divine 
ali'i are forbidden to have sexual relations with women of lower rank, 
and they are obliged — men and women — to remain virgin until mar- 
riage (NK, 2:88-89)." NK 2 is mentioned as a source for the custom of 
premarital virginity, not in support of its connection with ritual purity. 
If anyone makes this source say what it does not say, it is Chariot, who 
writes: "the authors go on to say that the emphasis on virginity in some 
Hawaiian legends is a result of missionary influence" (C. 108). What 
the authors actually say is as follows: "In Hawai'i's stories, missionary 
influenced writers-translators who first put them in written form may 
have injected their own bias for chaste heroines" (NK, 2:89). Thus 
Chariot transforms a "may" (that is, a hypothesis for which, inciden- 
tally, no evidence is given) into an "is." 2 This is an example of what he 
accuses me of doing: transforming a simple hypothesis into a proven 

I now address the second point, that is, the connection between vir- 
ginity and purity. Chariot claims "this virginity was maintained . . . 
for purely practical genealogical reasons, not for the maintenance of rit- 
ual purity" (C. 108). His basis for this important interpretive claim is 
Pukui, Haertig, and Lee, a modern text useful as a compilation of 

150 Pacific Studies, Vol. 10, No. 2— March 1987 

sources, but one whose interpretations should be used with caution. 3 In 
my book I have repeatedly attacked the very Western-modern view that 
Hawaiian rank was a "practical" or purely "political" matter and 
attempted to show that it is very much connected with ritual. I will not 
repeat these arguments here, but simply focus on the connection 
between virginity and ritual purity. In a footnote to my general discus- 
sion of pollution (V. 361 n. 12), I referred to a passage from the nine- 
teenth-century novel Laieikawai by Hale'ole, which supports the view 
that loss of virginity involves loss of purity (Beckwith 1919, 510-512). 
Chariot objects (C. 110) that this view is not traditional on the grounds 
that the heroine of that novel "receives in many ways a Victorian ideal- 
ization." Chariot forgets that Hale'ole was one of the great experts of 
Hawaiian tradition (see V. xxv) . His statement on virginity is at any rate 
confirmed by two texts that display no Victorian idealization. 

One text is the story of a chiefly woman who has lost her virginity and 
is therefore banished by her father. She is met by emissaries looking for a 
new wife for their king. This king has a virgin daughter who lives on top 
of an extremely taboo terrace (ka 'anuu kapu loa, Fornander 1916- 
1920, 4:545). This virgin daughter invites the heroine to sit by her on 
the platform, which, disliking the fact of her lost virginity, magically 
makes her slip. Later, the king's daughter brings the heroine to a 
bathing pool "which was also a very sacred place, those having lost their 
virginity, or who were defiled, were not allowed to bathe in it" (" 'a'ole 
e 'auau ka po'e i naha, a me ka po'e haumia," ibid.). When the girl 
attempts to climb the bank of the pond she is again mystically pushed 
back because she is not a virgin. Had a priest not discovered that she 
was of higher rank than the king himself, and therefore taboo to him, 
the girl would have been put to death for having defiled the sacred plat- 
form and pool (ibid.). 

The story demonstrates that loss of virginity implies loss of ritual 
purity, since it makes pure places mystically react against the deflow- 
ered woman. Furthermore, the text explicitly establishes a parallel 
between "deflowered persons" (po'e i naha) and "polluted persons" 
(po'e haumia). It seems to me, therefore, that far from being evidence 
"against" my thesis, as Chariot incongruously states (n. 3), this text 
proves that my thesis is correct. 

Another text illustrates the connection between virginity and ritual 
purity for a male. This is the story of Uweuwelekehau, a young man 
who "was always accompanied by his two gods, Kane and Kanaloa. His 
bringing up was surrounded by many restrictions; his house was sacred, 
people not being allowed to pass near it upon pain of certain death" 

Book Review Forum 151 

(Fornander 1916-1920, 5:194). The gods do not allow the man to have 
sexual intercourse with the woman he falls in love with, because he is 
"bound" by their mana ("ua pa'a i ka mana o Kane a me Kanaloa," 
ibid., 197). This text indicates that there is an incompatibility between 
the presence of divine mana (a ritual state) in the man and sexual inter- 
course, and therefore vindicates my position against Chariot's criticism. 
On the other hand, he is right in saying that my sources do not prove 
that chastity belts were used to preserve virginity before marriage: they 
only prove that chastity belts preserved fidelity within marriage. 

2. Chariot questions the validity of three examples that I give to illus- 
trate my point that mythical ali'i "are readily placed at the origin of cer- 
tain species, especially foods" (V. 146). He acknowledges that my third 
example "fits" but claims that my first example "could better be under- 
stood as a story of the gods; and the second . . . does not state that the 
persons involved are chiefs" (C. 109). Had Chariot departed from his 
usual practice (at least in this "review") of reading texts out of context, 
he would have discovered that his claim concerning the second example 
is false. Indeed the text to which I refer is but the continuation of 
another one, where it is explicitly said that the person (not persons!) 
involved — that is, Maikoha — who metamorphoses into the wauke 
(Broussonetia papyriphera plant), is the son of the ali'i Konikonia 
(Fornander 1916-1920, 5:269). As for my first example, I fail to see how 
it "could better be understood as a story of the gods," since in the text to 
which I refer Hinaaimalama and her siblings — from whom various spe- 
cies originate — descend from grandparents who are alternatively called 
ali'i (chiefs) and akua (gods) (ibid., 267). If anything, this myth proves 
my most general point — rejected by Chariot — that chiefs and gods are 
treated as interchangeable in many contexts. 

My next sentence in the book is also the target of Chariot's criticism. 
He claims that the chants in which chiefs are called various animals 
"cannot be used to demonstrate that the chiefs in question actually 
assumed the bodies named" (C. 109). My views are actually more subtle 
than Chariot represents, as is demonstrated by my statement: "It is as a 
shark, or by mystically controlling sharks, that the king often punishes 
transgressors and rebels" (V. 151). The statement that the king "is a 
shark" is further qualified on the same page — in the sentence quoted by 
Chariot — where I say that he is a shark, only in the sense that he can act 
through it and has a substantial relation to it, and so on. Chariot is 
wrong in interpreting this sentence as a claim that there is "an identity 
of the chief with the shark" (C. 109-110). But whatever the specific 
nature of the connection postulated between certain chiefs and certain 

152 Pacific Studies, Vol. 10, No. 2— March 1987 

animals, my basic point is that it is not conceived as a simple "meta- 
phor," but as a true "affinity" (V. 151). This is supported by the evi- 
dence, as Chariot must acknowledge. On the other hand, I find it hard 
to believe that he could seriously put forward the following argument 
against the view that ali'i can assume nonhuman bodies: "the chiefs 
honored in the chants are historic figures: Kalani'opu'u and Kameha- 
meha. Had they been able to assume nonhuman bodies, it would have 
been mentioned in the many historical accounts we have of them" (C. 
109). I thought that we were discussing beliefs in chiefly powers, not 
actual chiefly powers. But it seems that Chariot is prepared to believe in 
anything, even in the kind of objection that he is able to devise! 

3. Chariot finds fault with my statement that "sexual intercourse 
with inferiors is also polluting to superiors" (V. 91; see also 149). He 
expresses his surprise that I say this, since I have myself shown on page 
150 (and p. 372 n. 56) that such intercourse was extremely frequent. On 
the page to which he refers, I say that the rules concerning the sexuality 
of the ali'i reflect two contradictory requirements of chiefly rule: the 
preservation of the rank's purity by avoiding admixture with lower 
ranks, and the necessity of spreading life and of increasing the number 
of the chief's followers. I have also indicated, following Malo, that there 
is a partial solution to this contradiction: until an heir of the proper 
rank is born, sexual relations are restricted; after that, they become 
free. Chariot's criticism simply reflects his inability to see a statement as 
part of an argument: he sees two trees, but does not see that they are 
part of a wood. 

Chariot's attempts to disprove that two texts which I use support the 
thesis that sexual intercourse with inferiors is polluting for sacred ali'i 
are also misguided. The first one (Malo 1951, 70-71), he says, concerns 
"definitely a special case," since it refers to the ban on intercourse with 
"the kauwd or pariah class." I would not call it a "special case," but an 
"extreme case" since the kauwd are the extreme opposite of the ali'i 
from the point of view of rank. Chariot's claim that marriage with 
kauwd is presented as "bad for genealogical reasons — that is, one 
becomes declasse— rather than for ones of pollution" is typical of his 
superposition of modern Western ideas about "class" on the Hawaiian 
ideology of rank. As I have made abundantly clear, Hawaiian rank is 
measured by the degree of closeness to the gods from whom the ali'i 
descend. Those who are close to the gods are said to be pure, while the 
impure are separated from them. In Kamakau's words, "all those who 
were polluted were kept separated because the god desired only those 
who were clean and pure" (1964: 64). This implies that in Hawaii, as 

Book Review Forum 153 

elsewhere in Polynesia, the higher the rank, the greater the purity, and 
that loss of rank is loss of purity. Indeed I have quoted cases in which 
sacred ali'i, because they were polluted by their enemies, were reduced 
to very low status. It is because they are impure that the kauwa are low 
in status and therefore avoided as marriage partners by those who are 
higher. "Genealogical reason" is derivative, as is also shown by the fact 
that nonsexual forms of intercourse with the kauwa are sufficient to 
produce a loss in rank: "The houses of the slaves [kauwa] were tapu. No 
one not a slave could go there. If any one not a slave was seen there he 
became like an eating sore, a disgrace to his descendants" (Kepelino 
1932: 144). No mesalliance is involved here. 

To uphold his point that the taboo of intercourse with kauwa had 
nothing to do with pollution, Chariot (n. 4) is forced to do away with 
Malo's statement that genealogists try to ho'oma'ema'e or "cleanse" ali'i 
lines of connections with kauwa. Chariot claims that since ho'oma- 
'ema'e is not a "technical term" it need not have the sense of "cleansing" 
in such a context. What other sense, then? At any rate Chariot is in 
error in saying that ho'oma'ema'e is not used as a technical term: it is 
used by Malo, for instance, to refer to the purification of a woman after 
delivery (Malo 1951, 138-139), as I have noted on page 86 of my book. 
More importantly, Chariot forgets that Kamakau makes the same state- 
ment as Malo about cleansing ali'i rank, using the more common, and 
stronger ("technical," if he wills) term huikala, "purify," "cleanse": "By 
mixing here, mixing there, the blood of lords has become mixed with 
the blood of kauwa, and there is nothing that can cleanse it ('a'ole mea 
nana i huikala)" (Kamakau 1964: 9). 

Chariot also betrays an utter lack of understanding of the notion of 
"pollution" (n. 4). He says that the sentence "the kauwa class were 
regarded as a defilement and a stench" (Malo 1951: 71) is mistranslated. 
The original Hawaiian reads "ua kapa ia ka poe kauwa he palani, he 
hohono ke ano" which literally means "kauwa were called palani fish, 
a type that has a bad odor." Now to compare the kauwa to something 
that has bad odor is simply to call them "impure," "polluting." Indeed 
"impurity" and "pollution" are themselves physical metaphors of social 
states, as every anthropologist knows (Douglas 1966) and as I have 
myself noted (V. 85). Emerson's translation is therefore inaccurate but 
not conceptually misleading, as is confirmed by other texts. Thus 
Kamakau writes: "They [the kauwa] continued to hide their shameful 
blemish ('alina hilahila), but they could not wash out their tainted 
blood" (Kamakau 1964: 9). And Kepelino writes that kauwa were 
called hawahawa, a reduplication of hawa, which Pukui and Elbert's 

154 Pacific Studies, Vol. 10, No. 2— March 1987 

dictionary glosses as "defiled, unclean, filthy, daubed with excrement" 
(PE, 58). If this is not "pollution" what is it? (For these references and 
others, see V. 85, 164, 360 n. 4.) 

The second text discussed (Kamakau 1961, 128) says that the ali'i 
Kahahana became "degraded" and lost the "tabu of fire (Ahi), heat 
(Wela) and extra-ordinary heat (Hahana)" because he made love to 
"lesser chief esses." The connection between high status and purity is 
clearly stated in this text (V. 149); Chariot, however, says that "kapu 
often have particular rules, and one cannot generalize from one exam- 
ple." To this I will answer that my views on rank and its preservation by 
avoiding mixed unions do not rest on this single example and are not 
expressed in the single sentence of my book that Chariot quotes. If my 
critic wishes to demonstrate that Kahahana's case is exceptional, he 
should do so by proving (1) that the state of kapu is not for the high 
chiefs a religious state, that is, a state that involves a connection with 
the pure gods; and (2) that sexual relations, particularly with inferiors, 
are not usually conceived as inimical to a kapu state. 

I think that my entire book is an argument against the first point; as 
for the second, it is sufficient to remind the reader that all those who 
participated in temple rites — and were therefore in a state of kapu — 
were required to abstain from sexual relations on pain of death (Malo 
1951, 164). Sexual abstinence was also prescribed when planting, which 
involved contact with the gods (NK, 1:201). This shows that the incom- 
patibility between sexual relations and closeness to the gods (which 
implies a state of kapu) is such a general principle that the example of 
Kahahana cannot be considered exceptional. Therefore, Chariot's com- 
parison of the rule against sexual intercourse with inferiors attached to 
Kahahana's kapu with one kapus peculiar requirement to shield the 
"head from the sun's rays" is totally specious. 

4. Next, Chariot declares that "generalizations can be made only 
with caution from individual authors or works of literature" (C. 110). 
Example: the "extreme aspects of identification" between chiefs and 
their lands that I supposedly postulate can only be found in the chant of 
Ke'aulumoku "who, however important, represents a very personal, 
uncommon viewpoint" (C. 110). Whatever reason Chariot has to 
believe this, he does not share with us. Personally, I think that the view- 
point of an official bard at the court of two kings, and the half-brother 
of the wife of one of them (Fornander 1878-1880, 2:67, 157, 210) can- 
not be as idiosyncratic as Chariot claims. At any rate, any reader will 
see that I simply write "his [the king's] kingdom is assimilated with his 
body" (V. 146) and "the body of an ali'i is ritually interchangeable with 

Book Review Forum 155 

his land" (V. 152). Both statements imply (as is also made clear by my 
analysis of the luakini temple ritual) a relationship of symbolic equiva- 
lence or substitution, not "extreme aspects of identification," which only 
exist in Chariot's imagination. 

5. Chariot's final example is an example only of his inability to read a 
sentence as part of an argument. I am accused of having referred to 
Kahiolo and Elbert to support my view that chiefs "are characterized by 
immobility and inactivity" — a view which, in Chariot's opinion, would 
only be supported by "a pejorative remark by a foreigner, amply refuted 
by contemporary literature" (C. 111). (Actually I also refer to a 
plethora of texts, which mention that very high ranking ali'i do not 
walk but must be carried and are therefore in a prescribed state of inac- 
tivity [V. 147]; but Chariot conveniently forgets this fact.) His partial 
style of quotation hides the fact that I am referring to Kahiolo and 
Elbert only as sources for the mythical motive of the hero's inaction 
(Kahiolo 1978; Elbert 1956-1957). But, had Chariot taken the trouble 
of reading through my argument, he would have discovered that I com- 
pletely agree with Elbert in the view that this is only a provisional state 
of inaction, which inevitably turns into a state of action. Indeed, I 
write: "In mythology as in reality (cf. Beckwith 1940, 412-413) a time 
comes when passivity turns into an explosion of activity and the king 
reveals himself to be king precisely because he produces and acts" (V. 
149) . The two parts of the argument are separated because of a more 
general expository choice. I am arguing that ali'i are depicted as having 
apparently contradictory behaviors. Thus I first discuss one series of 
behaviors (inactivity, rank endogamy, invisibility, etc.), then the contra- 
dictory series (action, sexual intercourse with inferiors, visibility, etc.). 

In this case, as previously, Chariot, astonishingly, is not aware of the 
structure of my argument and is therefore under the impression that I 
am making erroneous or contradictory statements. Although I should 
have referred to Elbert in the second part of the argument too, no major 
fault is involved in quoting a source only for part of the information 
that it contains, unless this partial use has the purpose of making false 
accusations against the author, in the manner of Chariot. 

The remark by a foreigner that Chariot defines as "pejorative" is: 
"The highest point of etiquette among illustrious Hawaiians was, not to 
move" (cited in V. 147). As the reader can see, nothing pejorative is 
meant here. I have used this remark because it is confirmed by many 
other such observations on Polynesian chiefs. Much of Polynesian (and 
Hawaiian) aristocratic etiquette is concerned with bodily control as an 
outward symbol of form and plenitude, which are chiefly and divine 

156 Pacific Studies, Vol. 10, No. 2— March 1987 

attributes. I have often commented on the symbolic value of immobil- 
ity, both in the book and in papers (Valeri 1982, 1985), and I do not 
have to repeat myself here. But Chariot has a mysterious comment: 
"that Valeri takes immobility literally can be seen from 272, 336." I 
don't understand what he means by "literal" immobility. But let me 
make clear once again that for me physical immobility is a sign of a spir- 
itual or ritual state, as both passages of my book to which Chariot refers 
(V. 272, 336) demonstrate. The first reads: "This natural complex of 
relationships is perpetuated in a transformed state at the cultural level. 
The tree, immobile and attractive, appears as the complete, encompass- 
ing element. Thus it is transformed into the image of Kunuiakea, 'Ku of 
wide expanse,' the supreme form of Ku" (V. 272) . The second reference 
says: "as soon as the victim is captured, he is immobilized and set 
apart — a sign of his consecration" (V. 336) . 

These are, then, the best examples that Chariot can find of inaccurate 
references in a book of more than four hundred pages! And I have 
indeed shown that if there is a person guilty of inaccurate and selective 
use of reference, both to my text and to many Hawaiian texts, it is 
Chariot himself. No doubt any serious user of my book (as of any book) 
will want to check my references; but he will want even more to check 
Chariot's references. Readers will also want to reflect on this strange 
fact: All the above examples of "inaccurate" references come from a few 
pages of a chapter in which I criticize a thesis put forward by Chariot 
and attempt to prove the contrary thesis. 4 

The next section of Chariot's critique addresses the question of inter- 
pretation. Unfortunately, we are faced once again with the myopic and 
atomistic style of criticism with which Chariot has acquainted us. I am 
immediately accused of announcing my interpretation of a text instead 
of offering arguments in support of it. The first example given is once 
again an example only of Chariot's inability to keep focus on the reality 
of what I have written. He claims that my interpretation of a whole 
chant is based on one line only. But anyone can see that I do not claim to 
interpret the chant (V. 55). I am simply commenting on one of the met- 
aphors used in its third part. Likewise, my reference to the gods coming 
from Kahiki is limited to that specific context. Interpreting this refer- 
ence as a claim "that the gods must come from Kahiki every time they 
are invoked" (C. Ill) is a truly herculean feat of distortion. As Chariot 
himself is forced to acknowledge, I have often mentioned other abodes 
of the gods. 

Book Review Forum 157 

Next, Chariot questions my interpretation of the story of the origin of 
the Kanehekili cult. He writes: "Two problems for Valeri's interpreta- 
tion are that the story is about the priest, not the god, and that he dies, 
rather than being sacrificed" (C. 111). Chariot is apparently question- 
ing as unproven the equivalence of the god and his priest. Their equiva- 
lence, however, is suggested by the fact that they have the same name 
(Kanehekili) and even more by a statement at the end of the myth: 
"Those who had the head [of the priest], they worshipped it; and also 
his eyes, or his mouth; they were called the eye of the god, or mouth of 
the god, and so on" (Thrum 1909, 48-49, cited in V. 132). Through the 
priest's body, then, the god becomes accessible for worship. This fact 
also demonstrates that the priest's body has the same value as a sacrifi- 
cial victim's body; this is all I implied by saying that the priest (i.e., by 
definition, a sacrificer) is equated with a victim. I did not imply that 
the manner of his death was identical to that of an ordinary victim. 
Indeed, I have repeatedly shown in my book that the manner of death is 
far less important in Hawaiian sacrifice than the act of consecration, 
that is, of giving an object, animal, or person the status of mediator 
between god and worshipper. This is precisely what happens with the 
priest Kanehekili, since the pieces of his body mediate between the god 
Kanehekili and his worshippers. I may add now, however, that the mys- 
terious and sudden death of the priest in the god's temple suggests that 
the god has killed him in order to make the spreading of his cult possible 
by distributing pieces of the priest's body. The alternative proposed by 
Chariot ("the body of the priest could be efficacious without being con- 
sidered a sacrifice" C. Ill) is inacceptable because he does not tell us 
what would make it efficacious. I spare the reader my comments on 
Chariot's other "hypothesis," contained in his words "and so on." I will 
simply say that this example shows that Chariot has understood nothing 
of my interpretation of Hawaiian sacrifice. 

Let us now turn to the accusation that I can dismiss texts that do not 
fit my views. Two examples are given. The first concerns a cryptic refer- 
ence by Kamakau to "mana" as "property" of baits (V. 100). This refer- 
ence does not in the least contradict my basic definition of the term 
mana, which is "potency, to be potent," "efficacy, to be efficacious " 
"success, to be successful" — properties that baits should have (V. 98). 
Kamakau's text, however, is an instance of the occasional extension of 
the term mana outside the properly religious and ritual context. It is for 
this reason that I say that in this case mana is "banalized" (that is "made 
commonplace"), not because I want to dismiss Kamakau's text as irrele- 

158 Pacific Studies, Vol. 10, No. 2— March 1987 

vant. On the contrary, I mention it because it is relevant to my thesis 
that in Hawaii the notion of mana has a greater range of applications 
than in some other Polynesian cultures (cf . V. 100-101). 

The second example given by Chariot is puzzling. I have used certain 
lines of a chant for a limited purpose, more corroborative than demon- 
strative of an interpretation of mine. Chariot asserts but does not dem- 
onstrate that the rest of the chant contradicts my interpretation. Since 
he gives no reason to believe otherwise, I continue to consider the chant 
as "somewhat anomalous" because it contains lines that are not found in 
any other account of the rite in which it was chanted. At any rate I 
don't see anything in these lines that contradicts my interpretation of 
the rite. 

Chariot then addresses a methodological question that seems at first 
interesting: too few senses of a word versus too many senses. Unfortu- 
nately we are disappointed again. My interpretation of lines 613-614 of 
the Kumulipo chant is questioned on the grounds that I select only one 
meaning for kdne ("human male") and only one meaning for ki'i 
("image of a god"). According to Chariot, other meanings of the terms 
kdne and ki'i should be considered: hence it would be impossible to 
claim, as I do, that the god (Kane) is called after the man, and the man 
(Ki'i) is called after the god (V. 6). But I will stick to my interpretation 
because the context rules out any of the other meanings for kdne sug- 
gested by Chariot. "The word can be used for the human male, but for 
any other male as well: animal, vegetable, mineral, or god" (C. 112). 
The god Kdne is represented as neither mineral nor animal in the con- 
text of the Kumulipo, but as human, in fact all too human, since he is 
jealous of his wife and angry at Ki'i for his secret union with her (a 
union which, incidentally, means taking the god's place): 

She slept with Ki'i 

Kane suspected the first-born, became jealous 

suspected Ki'i and La'ila'i of a secret union [?] 

They pelted Kane with stones 

hurled a spear; he shouted aloud 

"This is fallen to my lot, for the younger [line]"[?] 

Kane was angry and jealous because he slept last with her 

his descendants would hence belong to the younger line 

the children of the elder would be lord (Beckwith 1951:106, 

lines 696-704). 

So much for animal, vegetable, and mineral! As for kVi, Chariot tells 
us: "A glance at the dictionary will show that Valeri is selecting only one 

Book Review Forum 159 

sense of ki'i' (C. 112). Let us glance at the dictionary to establish 
whether my choice is justified. KiH receives five groups of glosses in the 
dictionary of Pukui and Elbert: 

1. Image, statue, picture, doll, petroglyph; features, as of a 
face; plans, as for a house; carved, as end of an 'auamo 
pole . . . 

2. To fetch, procure, send for, go after, attack; to seek for sexual 

3. Hula step . . . 

4. Same as alani, sl tree. 

5. Gesture, as in hula. (PE, 136-137) . . . 

Glosses in groups 1, 3, and 5 are probably related, but they are 
obviously unrelated to groups 2 and 4, which are unrelated to each 
other. The signifier /cm, therefore, does not constitute one single polyse- 
mous word, but at least three different homophonous words. To suggest 
that all these meanings should be involved in the translation of the 
proper name Ki'i is to incur in the error denounced by Chariot himself 
of "using too many senses of a word" (C. 112). In fact, it would be a 
worse error: it would imply treating several homophonous words as one 
polysemous word! It seems to me that only a meaning in group 1 can be 
a candidate for translating Ki'i as used in this context of the Kumulipo. 
And among those meanings only the main one, "image" (that is, "image 
of a god," V. 6), makes contextual sense. As I have mentioned (V. 7), this 
translation receives further support from another episode related in the 
Kumulipo, where Wakea, the apical ancestor of all Hawaiians, conceals 
himself in the image of a god (ki'i) to seduce a divine woman. Here 
Wakea is like Ki'i relative to La'ila'i: hence it makes sense to say that the 
name Ki'i refers to a divine image in lines 696-698 (Beckwith 1951, 123, 
cf. 102-103 and the analogous myths in Valeri 1981). My translations of 
the names Ki'i and Kane in those lines are thus the most likely; Chariot 
cannot suggest any credible alternative, only a fan of dictionary mean- 
ings among which he does not choose. 

The next case is one of "too many senses." Chariot finds fault with my 
suggestion that the word lau is used in a magical formula for weakening 
the god Kamapua'a because of a deliberate word play on two signifi- 
cata: "numerous" and "seine." He comments: "A pig in a seine is an 
unusual image" (C. 112). Chariot is too literal-minded, which is not 
conducive to understanding the metaphors of a magical spell. The same 
literal-mindedness makes it impossible for him to understand my sug- 
gestion that the belt around the waist of the main temple image— called 

160 Pacific Studies, Vol. 10, No. 2— March 1987 

piko, "navel" — probably also evokes the two other references for piko: 
genitals and crown of the head (fontanel). Chariot objects: "A child can 
be circumcised, but how could the crown or fontanel be cut?" (ibid.). I 
never suggested that the symbolic "cutting of the piko" of the image was 
also a cutting of the fontanel. As a material act, the rite is only a cutting 
of the umbilical cord of the god, who is represented as a newborn child. 
But as a symbolic act, it probably constitutes the three piko that define 
every human: a connection with his consanguines (through the navel), a 
connection with his ancestors (through the fontanel), and a connection 
with his affines and descendants (through the genitals) . Indeed the lat- 
ter two connections are made possible by the first, that is by birth. The 
fact that the rite is symbolically overdetermined and that his various 
meanings are not reducible to the materiality of an act escapes Chariot 
who, as we shall see again and again, doesn't seem to understand the 
properties of symbolic thought. 

Next Chariot questions the connection that I tentatively suggest in a 
note between the two meanings of lele: "altar" and "messenger." But the 
two significata do have a common ground: their mediating role. 
Chariot gives no argument whatsoever against this connection. 

I am also accused of combining Kahiki as a place name with Kahiki 
as various cosmic points. But all these meanings constitute one single 
grouping in Pukui and Elbert's dictionary. I am simply attempting to 
make sense of a connection that clearly existed in the Hawaiian mind, as 
testified, for instance, by the text of Kamakau to which I refer in my 
book (V. 9). One passage of this text reads: "Here are some terms for the 
kukulu o ka lani, 'the borders of the sky,' or kukulu o Kahiki, 'borders of 
Kahiki.' These are what ka po'e kahiko called all lands beyond the 
Hawaiian archipelago — the lands beyond the circles of Kahiki-moe [the 
horizon] and Kahiki-ku [defined as "the (first) band of the firmament 
where it ascends upward" on p. 5]. These lands were called the lands of 
kukulu o Kahiki [kukulu is another word for "horizon," p. 5] or of 
kukulu o ka lani or of na paia ku a lani, the standing walls of heaven or 
of kumu lani" (Kamakau 1976, 6). Moreover, since Kamakau connects 
Kahiki with all lands beyond the horizon and the lowest zone of the 
celestial dome, we must conclude that these lands are indeed considered 
invisible from Hawaii. Indeed, contrary to Chariot's statement that I 
introduce in my discussion "the word 'invisible,' which does not appear 
in these texts" (C. 112), the word "invisible" appears in Kamakau's text, 
which speaks of the "invisible horizon" whose "only boundary is where 
it adjoins the solid walls of the sky" (Kamakau 1976, 5), where he 
situates the "lands of Kahiki-ku" (ibid., 6). Chariot completely misun- 

Book Review Forum 161 

derstands my definition of Kahiki as "invisible transcendent place." My 
references to the above texts should have made clear to him that I do not 
conceive of the lands of Kahiki as invisible in an absolute sense; they are 
invisible from Hawaii. Indeed I am well aware (since I have mentioned 
the fact several times) that mythical travelers sailed to Kahiki and set 
foot on land! This is another case where Chariot confuses his misconcep- 
tions with what I actually say. He also demonstrates that he is not a 
careful reader of the Hawaiian texts to which I refer, since he claims 
something completely untrue about at least one of them. 

Chariot also accuses me of dividing "words into parts to get more 
meanings" (C. 113). He can produce only one example for this: I am 
supposed to assert that "kauila wood" is to be interpreted as ka uila, 
"the lightning" (ibid.). I say nothing of the sort. I am referring to kauila 
as the name of a rite, and saying that this name may be "a metonymy 
for the feather gods, which are supported by a handle or pole made 
from kauila wood" and could also be associated "with lightning (ka 
uila), that is, a manifestation of the divine power in its luminous but 
violent (as befits the akua hulu manu [feather gods]) form" (V. 269). 

Chariot often claims that I formulate the correct methodology, but 
then do not follow it in practice. The problem is that he understands 
neither my methodological points nor what I do in practice. Two exam- 
ples of this double misunderstanding can be found on page 113 of his 
text. Chariot notes that I caution against "the temptation to arbitrarily 
construct a single account of the rites patched together from different 
sources" but on page 8 combine two texts to "connect Po and Kahiki" 
and, on page 331, different versions of a story. Chariot is confusing two 
different things. My methodological statement concerns patching to- 
gether different descriptions of whole rituals to construct a single 
descriptive statement; it does not concern, as should be obvious, the 
establishment of structural relations and equivalences at the level of 
interpretation. At any rate, nowhere on page 8 do I say that Po and 
Kahiki are connected. I say that they are two metaphors of "divine ori- 
gins" that differ in that Kahiki, contrary to Po, is one of "more concrete 
metaphors . . . with primarily spatial connotations" (V. 8). On the next 
page I state: "Po and to a certain extent Kahiki are metaphors for the 
undifferentiated state of the divine power, which is placed at the origin 
of the living universe" (V. 9) . If there is a connection between Kahiki 
and Po it is the connection between two different metaphors. With 
regard to page 331 of my book, I do not put together different versions 
of a story to produce an artificially constructed version. I am simply try- 
ing to uncover a system of relations common to all versions. 

162 Pacific Studies, Vol. 10, No. 2— March 1987 

Clearly, Chariot has no idea of the structural method, as is demon- 
strated by remarks found throughout his paper. For instance, he seems 
to believe that structuralism is a method based on finding "the raw and 
the cooked" everywhere. He even enlists me against Levi-Strauss, by 
misquoting one of my sentences: "Valeri states that 'the disjunction 
between the raw offering and the cooked offering' cannot be found in 
Hawaii (123)." What I actually say is: "It does not seem that the dis- 
junction between the raw offering and the cooked offering exists in the 
everyday appropriation of taro (cf. Kamakau 1976, 36-37); however, it 
exists during the celebration of the New Year and on other occasions 
when cultivators place their offerings of raw taro on the altar of their 
district" (V. 123). I hope that in his essays Chariot does not quote his 
Hawaiian sources in the way that he quotes my book! Note also that on 
pages 57-58 I show that offerings are differentiated into raw, roasted, 
broiled, and cremated. 

On pages 113-114 Chariot illustrates two of his claims: "Valeri's 
arguments from texts are often tenuous"; "Valeri's arguments are often 
very short." My arguments are tenuous and short only in Chariot's dis- 
torted rendering. The first of the two examples of "tenuousness" con- 
cerns my hypothesis that a correspondence exists between the state of 
the audience in a certain rite, a state described as 'olu'olu by K. Kama- 
kau (in V. 289), and the state — also described as o/u'o/u — induced in 
the god according to Malo (ibid.). Chariot objects that I translate 
'oluolu only as "affable," while the word has many glosses. I have 
already explained that only the context allows us to decide how to trans- 
late a word. K. Kamakau is describing the mood of the participants: 
this can hardly be described by such glosses as "refreshing," "soft," or 
"comfortable," to which Chariot contrasts my gloss "affable." The latter 
seems to me a good general term summarizing all glosses that are rele- 
vant in this context: "cool," "pleasant," "polite," "courteous." Moreover, 
it is not true that I translate 'oluolu by "affable" only: I also give it the 
meaning "cool" and "to soften" (V. 289). Thus I use all the relevant 
meanings of this word, while Chariot adds other meanings that are 
either redundant or irrelevant. 

Not happy with just distorting what I say, Chariot adds a strange 
argument: " 'Olu'olu is a common word, so it would be difficult to 
make the correspondence argument even if it were appearing in the 
same text" (C. 113). 'Olu'olu may be a common word, but it is not com- 
mon in the narratives of the luakini temple ritual. Hence the use of the 
same word to describe the state of both god and worshippers may be sig- 
nificant indeed, even if each state is described by a different source. 

Book Review Forum 163 

Furthermore, to object that the two states refer to "different points of 
the account of the ceremony" is to miss the point, which is, precisely, 
that one state results from the other, and is therefore preceded by it (V. 
307). Because I use the word "results," Chariot claims that I view the 
relationship between the two states as causal. Chariot could have 
spared himself this false claim if he had considered my statement (V. 
307-308) in the light of what I have said about the relationship between 
god and worshippers throughout my book: that it is dialectical and 
reciprocal, not causal. But I have also said that this reciprocal relation is 
often weighted toward the worshippers, as is made clear in several con- 
texts discussed in my book (V. 101-104, for instance), including the one 
misrepresented by Chariot (V. 307-308). It is because of this weighting 
that I say there that the state of the god "results" from that of the 

Another example given by Chariot is even worse. He claims that I 
give no reference for my various statements concerning the effect that 
dancing has on the development of the fetus of an ali'i. In reality I give 
several references to texts describing facts that lend themselves to my 
interpretation, particularly K. Kamakau (1919-1920, 2-4; because of 
an unfortunate typo, the printed text has "24" instead of "2-4"). K. Ka- 
makau says that the ali'i commands the people to dance in honor of his 
soon-to-be-born child and to compose and sing songs in his praise. These 
actions are said to have the power to ward off the negative effects of sor- 
cerers and angry gods on the fetus (ibid., 2). Does not this justify my 
interpretation that "dance is necessary to help develop the fetus of an 
ali'i and to ease his birth" and therefore "help engender" it? What other 
interpretation does Chariot propose? As for my statement that "dance 
contributes to affirming the reality of the ali'i's mana" I would claim 
that, given the common association between mana and growth (cf. V. 
96-97, 330-331), it is indirectly supported by K. Kamakau's text. It is, 
at any rate, supported by the facts mentioned on page 384, note 56. 

Chariot connects the above statements of mine and a hypothetical 
one derived from them (V. 219) with two other sentences in order to 
prove that I transform hypotheses into "confirmed facts" (C. 114). This 
"proof" is achieved by distorting what I say. The first sentence ("the 
engendering of Lonomakua, like that of any god" V. 219) has no con- 
nection with the hypothetical statement (also on V. 219) quoted by 
Chariot on the relationship between dancing and the engendering of the 
god Lonomakua. That no connection exists is made clear by the rest of 
the sentence, which Chariot omits: ". . .is represented as the growth of 
a human." This continuation indicates that the sentence refers, not to 

164 Pacific Studies, Vol. 10, No. 2— March 1987 

the engendering of the god through the dance, but to the fact that the 
process of construction and consecration of the god's image is repre- 
sented in ritual as the growth of a human (cutting the navel cord, gir- 
dling with loincloth) . 

Chariot has no more justification in using the next sentence that he 
quotes, since there the expression "is born of the feasting" is a purely 
metaphoric expression and one that implies a relationship between the 
production of the god and the feast as a whole, not with the dancing 
alone, contrary to what Chariot attempts to suggest. I also fail to see 
how Chariot can say that hypothesis on page 99 becomes a confirmed 
fact on page 101 when he leaves both unspecified. Such an allusive style 
of reference and criticism, which is abundantly used by Chariot, espe- 
cially in his footnotes, hardly corresponds to the conventional rules of 
scholarly debate. 

Chariot misreads my statement on page 273. It is not the statue made 
from the Haku 'ohi'a tree, but the tree itself that "is inseparable from 
the birds." Indeed in the section to which this sentence belongs, the 
name Haku 'ohi'a ("Lord 'ohi'a tree") refers to the tree or to the god in 
tree form, as is made clear by this passage: "In his 'wild' state the god is 
called by the name of the tree, Haku 'ohi'a" (V. 271). The immediate 
context of the sentence quoted by Chariot, in which I write that "the 
Haku 'ohi'a is fetched" (V. 273) from the forest, should leave no doubt 
of the fact that there Haku 'ohi'a refers to the god in tree form. 

As for the criticism of my allegedly "short" arguments, it can be dis- 
posed of in a very short time. I say that the relatives of the king are his 
"doubles" because they replicate his rank (V. 161, par. 3). Indeed the 
whole discussion in this chapter concerns the role of sacrifice in "rees- 
tablishing differentiation in a hierarchical system that, paradoxically, 
produces a certain coefficient of undifferentiation because of the over- 
lap of different principles" (V. 168). The statement on the identification 
of the transgressor with the king whose taboos he transgresses (V. 165) is 
not a "short argument," but simply the repetition of a thesis that has 
been previously argued in full (V. 92, 94). The statement on Atea (V. 
169), a Marquesan god, is a purely incidental remark, which is never- 
theless supported by my references (Tregear 1891; Williamson 1933). 
Chariot does not give any reason to believe that they are wrong. As for 
the statement on Kahoali'i (325), Chariot forgets that it begins with the 
expression, "It will be recalled that," which refers to a demonstration 
given on pages 260-262. I don't know who, apart from Chariot, would 
be in need of longer demonstrations for the next two statements of mine 
that he quotes. 

Book Review Forum 165 

Chariot also claims that Hawaiian pigs were smaller than modern 
Western ones (C. 114). This fact would make a speculation of mine con- 
cerning the age of a pig being sacrificed unlikely or at least unnecessary. 
European explorers, however, were impressed by the size of many 
Hawaiian pigs (Beaglehole 1967, 511 n. 1, 522, 1157, 1188). Samwell, 
for instance, notes that on the island of Hawai'i "there are great plenty 
of large hogs" (ibid., 1188). 

I leave to the reader to judge the value of Chariot's claim (C. 114) 
that I "eschew argument" because I use various stylistical conventions 
to link sentences or to announce hypotheses or speculations. It is Chariot 
himself — it seems to me — who eschews argument by quoting these 
expressions out of context and by failing to demonstrate that they are 
substitutes for argument. Such methods of criticism create much of the 
impression of the "review." Nor is this impression dispelled by the 
method (partial quotation and no argument) by which Chariot at- 
tempts to prove his view that I tend to use hypothetical points as if they 
were confirmed. I don't see how my statement — "This classification of 
the fish species is in large part hypothetical. It does in any case confirm 
the theory advanced . . ." (V. 26) — can be given by Chariot as an exam- 
ple of my supposed tendency. Anybody reading the rest of the sentence 
quoted by Chariot can see that "the theory advanced" concerns not the 
fish species, but "the principal 'aumakua species" (V. 26). The latter is 
well supported by the evidence offered by the principal 'aumakua spe- 
cies, and only receives further confirmation by the fact that it can be 
applied, hypothetically but not unreasonably, to the fish 'aumakua. 5 
Sadly, Chariot has misunderstood my argument. 

The supposed slippage from the hypothetical to the confirmed in the 
case of Kuka'ilimoku (V. 222) exists only in Chariot's imagination. Must 
I repeat constantly "or some other equivalent god" after Kuka'ilimoku? 
I have made abundantly clear that the latter name is only that of the 
main feather god in the island of Hawai'i, but that the differently 
named feather gods of other islands and dynasties are functionally 
equivalent to Kuka'ilimoku (V. 247). In my use, therefore, this name is 
just a shorthand for the entire type of these gods. As a result of such 
methods, Chariot's criticism often seems nothing more than strings of 
partial quotations, each treated as a valid argument. 

The results are no better when Chariot tries his hand at logic. He says 
that my entire book is based on a circular argument: my theory will be 
confirmed by the analysis of the ritual, but that analysis depends on my 
theory (C. 115). Chariot is probably under the impression that the her- 
meneutical circle is the same thing as a "circular argument," but even 

166 Pacific Studies, Vol. 10, No. 2— March 1987 

so, it is somewhat surprising that he can believe that I have first formu- 
lated my theory, and then simply proceeded to apply it to the ritual 
without making any change in the theory as a result of the analysis. The 
fact is that he is confusing an expository device (I first summarize the 
argument in schematic form, then proceed to illustrate and enrich it by 
the analysis of a concrete ritual action) with the actual steps of my 
research! Moreover it is surprising that he does not see that indeed my 
preliminary statements about gods, sacrifices, ali'i, and so on, are much 
enriched and made more complex the further I advance in the analysis, 
particularly in the analysis of the luakini temple ritual. 

As for the supposed "example" given by Chariot of this circularity, 
namely the relationship between the model of the hierarchy of the gods 
and the model of the hierarchy of the temples (C. 115), I will simply 
observe that 

1. Chariot does not take into consideration my statement that the 
first model is confirmed by the analysis of ritual, not only by that of the 
temple hierarchy; 

2. The evidence that I have offered on the temple hierarchy does not 
"cast the very idea of such a hierarchy in doubt"; it simply shows that 
several details of my model of this hierarchy are open to discussion. No 
discussion, however, is offered by Chariot: only a dogmatic statement. 
Others better qualified to judge the case than he have found my ideas 
interesting and worth incorporating (Kirch 1985, 258, 260, 262); 

3. It seems that Chariot is an extreme positivist: no conjectural model 
and no argument from coherence (cf. Dumezil 1948, 18) are admitted 
in his epistemological universe. Chariot himself hardly follows such 
strictures in his theses; no wonder, because otherwise he would not be 
able to formulate them. 

My "attitude toward [my] evidence" (C. 116), writes Chariot, is 
shown by my discussion of mana. I claim that we should not make too 
much of the rare occurrence of this word in the descriptions of the tem- 
ple ritual by K. Kamakau and Malo. My claim is based on the following 
chain of arguments: All occurrences in K. Kamakau's text are found in 
prayers, which suggests that mana was mostly used in prayers; but only 
very few of the many prayers that were uttered in the temple ritual are 
given by K. Kamakau and Malo; hence the rare occurrence of the word 
mana may only be due to the rare occurrence of prayers in these 
sources. Chariot objects that there are too few examples of prayers to be 
able to hypothesize, as I do, that the word mana "must have been 
included rather often" in prayers (V. 98). To this I will answer that even 
if the recorded prayers are few, they do establish a significant contrast 

Book Review Forum 167 

between "presence of mana" in prayers and absence of mana in prose 
descriptions. It is this contrast, and not the absolute number of availa- 
ble prayers, that is the basis of my hypothesis which, incidentally, plays 
a marginal role in my discussion of mana. The main role is played by 
another argument: ritual action itself is the best evidence on mana, 
since a key source (not simply "another source" as Chariot defines it), 
the Mooolelo Hawaii, says that the whole ritual is about the transmis- 
sion of mana (V. 98). Chariot cannot deny this point. Nevertheless he 
continues to believe, without argument, that since "verbal expressions" 
are the fundamental evidence, my "views are not supported by the 
texts." Must one assume, then, that for him the texts describing the 
luakini temple ritual and related rituals are not texts? 

Next Chariot argues: "Valeri's main thesis can fairly be said to depend 
on his interpretation of one section of the main temple ceremony" (C. 
117). He refers to the section that, in Malo's description, identifies the 
cutting of the navel cord and the girdling of the loincloth around the 
main temple image, with the image's transformation into an akua 
maoli, "true god" — that is to say, a true embodiment of the god (V. 314- 
315). Chariot immediately distorts my argument by leaving out the ref- 
erence to the girdling with a loincloth from his quotation of my thesis 
(C. 117). The sentence excised by Chariot is: "This birth rite and the 
rite for putting the loincloth on the god that follows it are identical to 
those performed for any male child to transform him fully into a social 
being." Only after having said this do I say that "the transformation of 
the god into the perfect type of the human male is thus completed." It is 
clear, therefore, that both rites, the cutting of the navel cord and gir- 
dling the statue with a loincloth, are involved in the transformation. 
Chariot, however, gives the impression that my thesis is based only on 
the navel-cutting rite. 

Against my view that the navel-cutting rite for the god symbolizes his 
"birth," 6 Chariot uses a strange argument. He says that both Malo's and 
Kelou Kamakau's descriptions of the similar rite performed for male 
infants of the ali'i rank show "that that ceremony could be separated 
from the birth" (C. 117) because it was performed in the temple, not in 
the house where the child was born. It seems that for Chariot this is in 
itself proof that the rite could have nothing to do with birth. This argu- 
ment can only be sustained by reducing birth to a mere biological fact, 
that is, the expulsion of the fetus. But this modern Western definition of 
birth was certainly not shared by Hawaiians. Indeed, I see nothing in 
Malo and Kelou Kamakau suggesting that the navel-cutting rite is not 
considered part of the process of bringing the child to life: they only 

168 Pacific Studies, Vol. 10, No. 2— March 1987 

imply that, in the case of male infants, this process cannot be left to 
women alone, but must be concluded by a ritual act performed by men 
in the presence of the gods. The small spatial and temporal hiatus 
between the parturition and the navel cutting cannot therefore be 
seriously used as an argument against my interpretation. 

Nor does the use of the navel-cutting rite as an image "to express the 
beginning of something" (C. 118) prove anything against my interpreta- 
tion 7 either, since the rite can function as an image of beginning pre- 
cisely because it evokes birth, which is the most compelling image of 
beginning. However metaphoric, then, the cutting of the navel cord of 
the god evokes the idea of birth; contrary to what Chariot claims, I do 
not go beyond the evidence in claiming this. Moreover, one should con- 
sider the navel-cutting rite in its syntagmatic context. This context 
shows that the images of gestation and birth permeate the ritual. The 
statue of the god is often treated as a fetus or as a baby in the rites pre- 
ceding the cutting of its "navel cord." Furthermore, the rite is followed 
by the girdling of the loincloth on the statue. This sequence does suggest 
a passage from birth to social adulthood: therefore the cutting of the 
navel cord cannot be considered as a mere beginning; it is a much more 
concrete image. Once again, we see that Chariot commits the capital 
methodological sin of interpreting a fact out of context. 

Chariot also states that "the primary object of the ceremony under 
discussion is the statue, and some discussion is necessary on which cere- 
monial points apply to it and which to the god, however one conceives 
of the relation between the two" (C. 118). On the basis of Malo's (1951, 
171) statement that when the sequence of rites concerning the image is 
concluded, the image becomes an akua maoli, "a real god," I would say 
that Chariot's query is meaningless, since obviously the statement 
implies that by acting on the image one acts on the god, that to the visi- 
ble process of the ritual corresponds the invisible process of the god. The 
idea of such correspondence is well known from other parts of Polynesia 
as well (cf. Firth 1970). 

Chariot's argument that I go "beyond the evidence in describing the 
section of the ceremony as the birth of the god" can fairly be said to 
have been disposed of. But what about his other claim: "His [Valeri's] 
characterization of that god as a man is derived wholly from his theory 
and has no basis whatsoever in the text" (C. 118)? Reading this sen- 
tence, I wonder what qualifies as textual evidence with Chariot. From 
all his arguments it appears that only explicit statements qualify as 
such. If one can find in a text a sentence that supports an interpretation 
by explicitly saying that a thing is indeed as the interpretation claims it 

Book Review Forum 169 

to be, the text can be said to support the interpretation; if not, it does 
not. Obviously, Chariot confuses one level of the text, and one mode in 
which it provides evidence, with the notions of text and textual evidence 
in general. 

If he means that my "characterization of that god as a man" is not 
supported by the text because the text does not contain the sentence 
"that god is a man," I agree with him. Indeed, it would be surprising if 
such statement existed. As a matter of fact, its existence would contra- 
dict my theory, which is based on the assumption that the human char- 
acter of the god is recognized only in mystified form by Hawaiian con- 
sciousness (otherwise it would not be religious, V. 345-346) . I therefore 
expect to find only clues to that identification in the textual material. 
But these clues are quite clear. If a god is represented in ritual — how- 
ever metaphorically — as being generated, born, and given a loincloth, 
am I wrong to say that he is represented as human? If, moreover, the 
statue representing the god is the icon of a perfectly developed man, 
both physically and socially (in that he recalls the highest ranking ali'i), 
can I not say with some justification that he is "the perfect type of the 
human male"? Chariot is blinded by his literalism. He disregards the 
fact that a text is a complex entity that communicates in different forms 
and at different levels, not only at the propositional one. I would main- 
tain, therefore, that he misses a whole dimension of the meaning of the 
Hawaiian texts to which we both refer. 

Finally, Chariot addresses "one further difficulty for Valeri's theory": 
only Malo describes the "birth" rite. Chariot comments: "That is, what 
should be the most important ceremony of the whole sequence is 
replaced by a different one in two of the three sources" (C. 118). This 
objection disregards the fact that I do not give the same importance to 
all sources, but consider Malo as the most valuable. More importantly, 
it reflects Chariot's initial distortion of my position. As I have indicated 
above, the final transformation of the god, and therefore the most 
important ceremony of the whole sequence, is produced in Malo's 
account not simply by the "birth" rite, but by the sequence "birth" rite 
and girdling of the image with a loincloth. Indeed, it is the latter rite 
that is the fundamental one, since it marks the god's accession to full 
"manhood," the accomplishment of his development. Now this rite is 
also mentioned by the second most important source, K. Kamakau. 8 

Another unacceptable claim made by Chariot is that my thesis (V. 
315-317) on the paradigmatical relation between the makii lohelohe 
rite in K. Kamakau's account and the "birth" rite in Malo's account is 
invalidated by the lack of "any hint of birth" in the first rite. But the 

170 Pacific Studies, Vol. 10, No. 2— March 1987 

"paradigmatic relation" between the rites only implies that they are 
substitutable in the same context — not that they have the element of 
birth in common. The fact that this element is not mentioned in K. Ka- 
makau's text does not invalidate my thesis, which is based on the entire 
sequence describing the development of the god, a sequence common, 
with minor variations, to all the sources. 

But my analysis also makes clear that the birth rite in Malo and the 
maki'i lohelohe rite in K. Kamakau have one important element in 
common: the presence of "cords" that can all be considered as "navel 
cords" of sorts (and therefore metaphoric extensions of the birth image). 
In the birth rite, the image has a "navel cord"; in the maki'i lohelohe 
rite, the tower which, like the statue, is a device for rendering the god 
present, has four cords that are placed each at one of its corners. That 
these cords are like navel cords is a hypothesis, but one that makes sense 
(V. 316). The connection between the two rites illustrates my point that 
the texts describing the luakini ritual describe different practices, but at 
the same time share certain sense relationships (V. 317). Chariot's criti- 
cism reflects the arbitrary belief that material differences necessarily 
imply differences of meaning. Finally, let me say that it is simply by 
willfully ignoring the scrupulously maintained difference between the 
sections where I summarize the evidence and those where I give my 
interpretations, that Chariot can attempt to apply to my statements the 
criticism that I level against the statements of those who do not make 
this distinction: "It is not clear whether they are produced by the infor- 
mants or by the authors" (V. 51). 

After all the belaboring of details, for the most part without much 
importance, it is a relief to turn to what seems to be a more serious dis- 
cussion: that of my "theoretical orientation" (C. 119). Unfortunately, 
we are quickly disappointed. Chariot begins by piecing together a few 
statements from my book that refer to some sources of my theoretical 
inspiration: Hegel, Feuerbach, and Durkheim. He immediately dis- 
plays his talent for misunderstanding and distortion. For instance, he 
says that I take from Feuerbach the idea that Hawaiian religion is 
anthropomorphic. That Feuerbach wrote about Hawaiian religion is 
news to me. Nor have I made the mistake of associating the age-old the- 
ory that religion is anthropomorphic with Feuerbach alone. Chariot is 
confusing this theory with Feuerbach's thesis that man's consciousness 
of himself as species-being is reflected in his gods. He is misquoting from 
my page xi, where I say that "my argument has a certain Feuerbachian 
ring." Certainly, it is not the extremely common thesis of the anthropo- 

Book Review Forum 171 

morphic character of religious representations that has a Feuerbachian 
ring! I am also accused of having taken from Feuerbach the idea "that 
the state is a projection, so to speak, of human nature or essence" (C. 
119). However, I do not speak of "state" anywhere in my text. 

Quite novel, and indeed unprecedented in the annals of scholarship, 
is the method Chariot uses to "demonstrate" that a "central idea" for 
my interpretation of Hawaiian ritual and sculpture is inspired by a 
Brahmanical saying. The method consists in leaving unmentioned the 
fact that I quote the saying in a passing remark (V. 358 n. 65) on the sac- 
rificial gift in general, not on Hawaiian sacrifice specifically. Although 
he must acknowledge at one point that I myself often stress the differ- 
ences between Brahmanical ideas and Hawaiian ones, Chariot claims 
that I have derived my interpretations from Indian conceptions. His 
"proofs" for such statements remind me of the "proofs" adduced by an 
anonymous writer of about the year 200 to "demonstrate" that Homer 
borrowed from Moses: "Among his many 'proofs' were the 'borrowing' 
of the opening of Genesis for one bit of the description of the shield of 
Achilles in the Iliad, the portrayal of the Garden of Eden in the guise of 
the garden of King Alcinous in Book VII of the Odyssey; and Homer's 
referring to the corpse of Hector as 'senseless clay,' copied from 'Dust 
thou art and to dust thou shalt return' " (Finley 1977, 167). 

Other statements by Chariot remind me more of the art of collage 
than of the rules of serious scholarly argument. An example of this art 
(perhaps in its surrealist form) is his list of "terms from Western philoso- 
phy and religion" that he says I frequently use in my book "without 
explicit justification." The justification, when it is needed, is given by 
the analysis itself. For instance, "particular" and "universal" are used 
(V. 270) to describe the passage of the temple image from a "particular" 
(that is, an individual phenomenon) to a "universal" (a term applying to 
ail individual phenomena included in it) . What concrete objection does 
Chariot have to the use of such terms? Does he wish to claim that 
Hawaiians do not think in terms of particulars and universals, like all 
other humans? Don't they make a distinction between, say, "a man" 
and "man"? Note also that a majority of the terms mentioned by 
Chariot are used by me in quotation marks, to signify either that they 
are taken in a special sense or that they are used by default, because no 
better term is available. Such is the case, for instance, of "supernatu- 
ral," which I use only in a couple of cases, contrary to what Chariot 
implies (see below, p. 188). 

Chariot is not averse to stating plain falsehoods either, as when he 
claims that I use "substantialist" to "translate Hawaiian." What I write 

172 Pacific Studies, Vol. 10, No. 2— March 1987 

is "I would give the word la a a [the indefinite article was omitted 
because of a typo] more 'substantialist' meaning" (V. 363 n. 3). I am 
afraid that such a style of criticism reflects more on the critic than on 
the person criticized. 

Other terms to which Chariot objects without argument are "sacred," 
"profane," and "substance," which are extremely common in anthropo- 
logical discourse, where they do not have the philosophical or religious 
connotations that Chariot attaches to them. Analogously, in my (rare) 
use, the term "creation" is only a generic term referring to the multiple 
modes of the gods' productive activity (see below, p. 196). Note also 
that some of the terms whose use Chariot criticizes are in fact glosses 
taken from translations of Hawaiian texts or from dictionaries. For 
instance, he gives the impression that I use the word "miracles" fre- 
quently and as a matter of course. In fact, I used it only as a gloss given 
by Pukui and Elbert (PE, 53) of the expression hana mana (V. 
324 n. 26). 

More importantly, Chariot ignores the fact that anthropological 
interpretation is first and foremost translation; we cannot escape 
translating the ideological terms of another language into our own ideo- 
logical terms. Chariot does not escape this predicament either. In fact, 
as I have already indicated, he is prone to borrowing the tritest terms 
from the arsenal of Western common sense. We have already encoun- 
tered "practical," "class," "declasse"; we shall soon encounter "religion 
and natural science," and many others. 

Ironically, Chariot does not hesitate to accuse me of reading "classic 
Western views" into Hawaiian thinking. The list of such "readings" is 
simply ludicrous. I am supposed to think that in Hawaiian thought 
humans are "separate from nature," although I say just the contrary. 
Evidently Chariot has not read my interpretation of firstfruits sacrifices 
(V. 76-78), where I claim that since no difference is felt to exist by itself 
between humans and nature, nature could not be appropriated by man 
if ritual did not create some difference, although pretending all the time 
to reaffirm the absence of any difference (cf. V. 34, 359 n. 75). Further- 
more, all my interpretation of Hawaiian ritual is based on the principle 
that in various aspects of nature Hawaiians found aspects of themselves 
and of their social and mental life, that in Hawaii nature is naturalized 
man (Levi-Strauss' "l'homme naturalise") and man is humanized na- 
ture. Hence, saying that I claim that humans are "separate from 
nature" is tantamount to showing that one has not even understood the 
most basic thesis of my book. 

Even more ludicrous is Chariot's thesis that I separate "religion from 

Book Review Forum 173 

natural science," a statement he supports by referring to page 35 of my 
book. The reader who takes the trouble to read the page will discover 
that there I simply criticize Horton's "intellectualist" theory, which pos- 
tulates strong similarities between "primitive religion" and modern sci- 
ence (n.b. "science," not knowledge). I suggest that this theory does not 
apply to Hawaiian religion. Chariot understands this as reading into 
Hawaiian thinking a Western distinction between "religion" and "natu- 
ral science"! I shall discuss below other supposed instances of Western- 
influenced interpretations mentioned by Chariot. 

My criticism of Horton's intellectualist theory of religion should show 
Chariot's statement that I am an extreme intellectualist to be simplistic 
at best. But it is simply absurd to view my various statements on sym- 
bolic identity or "substitutability" as proofs of my supposed "extreme 
intellectualism." Symbolic substitution and its correlate, symbolic iden- 
tity, have nothing to do with "identity" in the sense that this term has in 
formal logic, contrary to what Chariot believes (C. 120). As I have 
made abundantly clear, the idea of substitution — in sacrifice, for exam- 
ple — postulates the symbolic identity between a thing or person that 
stands for another, at the same time that it presupposes their actual dif- 
ference. 9 It is precisely on this combination of postulated identity and 
actual difference that the efficacy of ritual, and more generally of sym- 
bolism (cf. Valeri 1981), rests. The identity of the king and his adversary 
or transgressor of his taboo is not an absolute identity: it is identity rela- 
tive to a certain quality that is highlighted — for instance identity rela- 
tive to rank, kapu, powers, ambitions, and so on. What normally 
should be defined as a simple similarity becomes something more than 
that, however, because it is correlated with a symbolic and even psycho- 
logical process of identification. In other words, the "other" becomes a 
"double" (cf. Girard 1972; Rank 1925; Vernant 1974 2:65-66). This 
seems to me to imply that a person and his double relate in a mode 
already oriented by the rules of ritual and more specifically of sacrifice. 
In other words, the transgressors or adversaries of the king are already 
seen as his sacrificial substitutes, for reasons that I explain in chapter 5. 

In sum, it is because he denies the essentially ritual nature of the 
hierarchical system and wants to see it in exclusively Western, "prac- 
tico-genealogical" terms, that Chariot completely misunderstands my 
use of the terms "identical" and "substitutable." To my use of such terms 
—motivated by a symbolic "logic" that eludes him— Chariot opposes 
his Western common sense, saying that such identities "would be natu- 
rally impossible" (C. 121, italics added). 

Chariot also attacks my argument that twinship is logically related to 

174 Pacific Studies, Vol. 10, No. 2— March 1987 

the idea of royalty with his usual argument that I "impose" my own 
terms on Hawaiian culture. He is unable, however, to demonstrate that 
my use of such terms is arbitrary, nor does he propose an alternative 
explanation that would demonstrably be closer to Hawaiian ideas. Such 
criticism is very easy but does not cut very deep. In a passage of my book 
I write that "circular things and things capable of circular movement 
are often considered divine, especially if they are powerful and distant, 
such as the stars or the moon" (V. 88-89) . Chariot's objection to this — 
"many nonround objects were considered akucT — is surprising, to say 
the least. Where and when have I said that only round objects were 
considered akua? And what can we make of this "objection" to my point 
that the moon and the stars are used as metaphors of gods because, as 
circular objects capable of circular movements, they evoke autonomy 
and self-sufficiency: "A traditional expression for feminine beauty — 
. . . , 'the face is like a moon' — does not evoke autonomy and self-suffi- 
ciency" (C. 121)? Evidently Chariot thinks that comparing the moon to 
the gods in one respect is incompatible with comparing it to feminine 
beauty in a completely different respect (shininess and roundness of 
face, etc.)! 

His next observation on my supposed contradictions in the characteri- 
zation of chiefs and gods is once more due to his failure to grasp the 
structure of my argument, as already indicated. Indeed, Chariot's con- 
stant repetition of the same misunderstanding is most tiresome. As for 
his claim that a chant that I mention (V. 371 n. 51) as an example of the 
ali'i's tendency to express themselves "metaphorically, poetically" on 
certain occasions "contains not a single metaphor" (C. 122), it is simply 
untrue. The chant contains one clear case of metaphor ("eaten by deep 
sorrow") and possibly others (for instance, "skies" and "mountains" 
may refer metaphorically to chiefs) . At any rate, I mention the chant as 
an example of poetical expression in general, not simply as an example 
of that particular form of poetical expression which is metaphor 
(V. 148). 

Nowhere do I say that only the poetic genre was used in communica- 
tion among chiefs, nor that— most absurd of all — "Hawaiian chiefs and 
chief esses had rank-related difficulty verbalizing" (C. 122). All I imply 
is that while, as Chariot says, "all classes of Hawaiian society used 
poetry," this use was much more important and elaborate among chiefs, 
as testified by Tl, among others, in the passage to which I refer (V. 
148). To say, as Chariot does, that there is no evidence for my statement 
that high-ranking nobles avoid situations of laughter is to ignore, for 
instance, that they cannot participate in crucial rites marked by laugh- 

Book Review Forum 175 

ter (V. 284). It is also to ignore a text that I mention (V. 276). On the 
other hand, Chariot gives no source for his claim that Kamehameha (a 
chief who, incidentally, did not have the highest rank and lived in a 
period of change) made his courtiers laugh. The few mentions of laugh- 
ter that I have found in the narratives concerning the kings of the past 
refer to an anomalous king and to private or at any rate not ritually 
charged situations. 

Finally, anyone who has taken the trouble to glance through the most 
elementary introductions to Buddhism will find it difficult to take 
seriously Chariot's accusation that I extrapolate an "arhat-\ike ideal" 
from the evidence on Hawaiian ali'i. How, indeed, can my account of 
Hawaiian ali'i be compared with "the arhat ideal, that of the human 
being who, by strenuous effort, acquires Enlightenment" (Humphreys 
1985, 49)? Where have I suggested that Enlightenment was a religious 
concept of the Hawaiians? And how can Chariot dream of comparing 
the hereditary rank of the ali'i with a state reached individually "by 
strenuous effort"? 

Chariot announces next that my idea that Hawaiian religion and 
society constitute a system is derived from Hegel. But it is not necessary 
to know anything about Hegel, just to read my book, both for its ana- 
lytic practice and explicit methodological declarations, to realize that 
my notion of system derives from Structuralism and particularly from 
Saussure. Chariot seems to have read only bits and pieces of my book 10 
and seems not to have read Hegel at all. He accuses me of believing that 
Hawaiian religious phenomena can be reduced to "a single, unified sys- 
tem" (C. 122-123) and of ignoring the "evidence of disunity." These 
accusations are strange, to say the least, since Chariot himself must 
acknowledge that I mention contradictions, conflicts, and differences 
from island to island. 

Indeed, I have elicited information hitherto neglected that seems to 
prove that the hierarchy of the major gods on Kaua'i (and even on 
Maui) was different from that on the island of Hawai'i (V. 185, 335). 
Chariot also mentions my view that the systematization of Hawaiian 
religious beliefs and practices is due to priestly and chiefly influence, 
although he fails to add that I stress the constant tension between spon- 
taneous, unsystematic creation from below and systematization from 
above (V. 36) . This tension contributes to the historicity of the Hawai- 
ian religious system, which I have never denied. I am all for considering 
Hawaiian religion historically. Indeed, I have written a special essay 
(Valeri 1982) based on an actually documented case of historical 
change, to which I have repeatedly referred in my book. But to pretend 

176 Pacific Studies, Vol. 10, No. 2— March 1987 

to study the "historical development" of the priestly system in the 
absence of positive historical documents, as Chariot suggests that I 
should have done, is to confuse the writing of history with the writing of 
science fiction. Writing a synchronic study of Hawaiian religion at the 
end of the eighteenth century, as it is documented by the sources, means 
to be more historically minded than to indulge in the gratuitous pseudo- 
historical fantasies of Chariot (C. 123-124, C. n. 33 in fine). To con- 
struct arbitrary diachronic sequences does not make one a historian. 

Chariot professes to be very surprised when, having given much evi- 
dence of conflict, diversity, and having warned that my analysis con- 
cerns the system of the island of Hawai'i (which is essentially the one 
documented in the sources), I claim: "I have attempted to give a coher- 
ent picture of the Hawaiian ideological system by considering all availa- 
ble information" (V. 191). But Chariot's surprise is due only to his mis- 
understanding of my use of the notion of system in this statement. I use 
"system" here in its structuralist sense: an abstract group of relations 
underlying a variety of concrete configurations, even conflicting ones. 
The system that I have attempted to reconstruct is neither the "priestly" 
system nor the "popular" system, nor is it the system of the island of 
Hawai'i alone; it is a set of relations underlying them all, and such that 
it does not deny, but makes intelligible their differences. This is pre- 
cisely the structuralist notion of system. But having mistaken my use of 
this notion for the "Hegelian one," and having on top of that confused 
the latter with the crassest empiricist use of "system," Chariot claims 
that I contradict myself or that I want to reduce all religious views and 
practices to the priestly system of the island of Hawai'i pure and simple. 
I repeat that since most of our documents do indeed refer to that island, 
we have no choice but to use them; at the same time, we can abstract 
from the system of the island of Hawai'i more general principles that 
are not necessarily in conflict with whatever facts we know about other 
islands or other views. I have not claimed anything else. But not only 
does Chariot misunderstand my most general point, he also distorts and 
falsifies the evidence in a hopeless attempt to prove that I contradict it. 

One example concerns his reading of some of my statements about 
priestesses, prophets, and sorcerers. He accuses me of calling these reli- 
gious figures "marginal" because, he says, they do not fit my system (C. 
123). He is simply disregarding the fact that this "marginality" is an 
indigenous evaluation. Malo, for instance, writes that the prophets 
(kaula) "were a very eccentric class of people. They lived apart in desert 
places, and did not associate with people or fraternize with any one" 
(Malo 1951, 114, cited in V. 138). "Marginal" means "to be at the mar- 

Book Review Forum 177 

gins." Are not asocial people who live apart in desert places marginal? 
Moreover Chariot confuses my structuralist use of the notion of "mar- 
ginality" (derived from, among others, Victor Turner, Edmund Leach, 
and Mary Douglas), which is value-free and implies that what is mar- 
ginal is powerful and therefore important, with his everyday use, which 
implies lack of value and importance. 

It is unforgivable that Chariot quotes page 328 of my book as evi- 
dence for the alleged fact that priestesses and prophets "take part in the 
temple ritual" and cannot therefore be considered marginal. The 
sources that I quote there do not mention priestesses at all — only high- 
ranking female ali'i who do not officiate and who, therefore, have no 
priestly function. Furthermore, they participate in a rite that takes 
place outside the temple proper (V. 237) and at the conclusion of the 
main ritual (V. 327-328). It is, therefore, structurally marginal, espe- 
cially since, as I have made clear, it marks the passage of men from the 
sacred world of the temple to the profane world of everyday life (V. 
326). As for the prophets, they come on the scene even after the female 
ali'i, and they worship the goddesses (not the gods, who are the object of 
the temple worship), imploring them thus: "Make the ali'i treat us well 
when we are in their presence, and see to it that we are granted forgive- 
ness (kala) when we ask for it" (V. 328) . This prayer indicates that the 
prophets are not simply marginal relative to the chiefly-centered temple 
ritual, 11 but are actually in a state of tension and potential conflict with 
the ali'i, since they have to implore the goddess to intercede with the 
ali'i on their behalf. Indeed, goddesses mediate between the temple rit- 
ual and the religious practices of the prophets, but are not part of that 
ritual in a strict sense. By claiming the contrary (C. 126), Chariot makes 
an incorrect statement and has the gall to give as supporting reference a 
text of mine that denies it flatly. 

Moreover, the sorcerers (more exactly, the black sorcerers) are not the 
"target" of my "polemic" (C. 123), but of that of most Hawaiian texts 
known to me. Kamakau, for instance, refers to black sorcery as "these 
evil ways of killing men" (Kamakau 1964, 137; cf. V. 138). Evidence of 
the sorcerers' categorical marginality is given by the texts that I quote 
on pages 138 and 370 (n. 31). Chariot refers to these pages but, surpris- 
ingly, he seems to ignore their content. Even more surprising are the 
other references (pp. 183, 185, 247-248, 380 n. 9) he uses as evidence 
that sorcerers are not marginal. At these pages I refer to the sorcery gods 
of the king, who form a very special category, including forms of the 
major gods Ku, Kane, and probably Kanaloa as well. The king himself 
is the priest of many of these sorcery gods, who are meant to counteract 

178 Pacific Studies, Vol. 10, No. 2— March 1987 

the practices of black sorcerers (cf. Valeri 1982). No ordinary sorcerers 
are involved in these rites, contrary to what Chariot implies. 

Chariot's criticism of what he calls my notion of the pantheon is sim- 
ply appalling. Each sentence contains so many errors and absurdities 
that to expose them all would require a separate rejoinder. I shall there- 
fore concentrate on the essentials. "Valeri's discussion of the 'pantheon' 
shows that (1) he wants to make it all-encompassing for Hawaiian reli- 
gion as a whole, and (2) that he wants it to be coherent" (C. 124). From 
this first sentence and what follows, it appears that Chariot identifies 
"pantheon" with the "system" of the four major gods and their particu- 
larizations. This is a very curious interpretation. Pantheon simply 
means "the assemblage of all the gods; the deities of a people collec- 
tively" (OED). It is with this meaning that I use this word, as other 
Polynesianists — for instance Firth (1970, 85) or Johansen (1954, 218) — 
have done before me. I do not limit its reference to the system of four 
gods and their particularizations, as should have been evident to 
Chariot had he read, at the very least, the subtitles of the section "The 
Pantheon" in my first chapter. These subtitles are: "The Major Gods 
and Their Particularizations," "Goddesses, Akua Wahine," "The Akua 
'Aumakua" "The Akua 'Unihipili." In other words, "pantheon" simply 
means "all the gods"; whether or not these gods are systematically 
related is irrelevant to my use of this term. For this reason, the state- 
ment that I make the pantheon "all-encompassing for Hawaiian reli- 
gion as a whole" is either tautologous (the set of all the gods includes all 
the gods!) or incorrect (since I do not refer to the four major gods by the 
word "pantheon"). 

Chariot also accuses me of basing my view that the "pantheon" is 
"coherent" "on systematizing nineteenth-century sources, especially 
Kamakau, and on the conventional idea of 'the four great gods of 
Polynesia' " (C. 125). This statement contains an error of fact and a 
grave misunderstanding. The grave misunderstanding consists in assim- 
ilating my hypothesis (this is indeed the status that it has in my book, cf . 
V. 110) that minor deities are encompassed by the major ones with 
Kamakau's thesis that "subordinate gods are produced by a 'segmenta- 
tion' (mahaeana) of the major gods" (V. 14). Kamakau's thesis does not 
correspond to my view, as should be evident even from the sentence 
quoted by Chariot to illustrate his claim (V. 36). There I refer to the 
already mentioned tension between the spontaneous proliferation of 
gods and the priestly attempts at systematizing them— a position that 
amounts to denying that Kamakau's segmentation model applies to all 
gods. While I consider Kamakau's model as one interesting conceptual- 

Book Review Forum 179 

ization of the relation between major and minor gods (cf. V. 14), my 
own view of this relation is more complex than his. Readers of my book 
will know that minor gods can be considered as encompassed by the 
major ones principally because their cults are hierarchically linked. For 
instance, I have given evidence proving that minor gods could not be 
worshipped until the worship of the major gods was completed in the 
royal temples (V. 187-188). Furthermore, the gods of the subordinates 
of a king or major ali'i (even their 'aumakua or "family gods") and those 
of their lands had to participate in the rites of the major gods where 
they were relegitimated and reconstituted (V. 263-270, 281-282, 290, 
etc.). In sum, my basic argument is that the hierarchy of the gods must 
be understood through the hierarchy of their rites and of the places and 
times of their performance. 

Because Chariot refuses to face the fact that the relations between 
Hawaiian gods cannot be interpreted independently of their relations in 
ritual, he fails to understand my Dumontian (cf . V. xv) use of the notion 
of "encompassment" (Dumont 1966). The use of this notion does not 
imply the view that all minor gods are forms of the major ones 
(although many are), but that the major gods are presupposed by the 
minor ones, 12 at least in the priestly theory that is predominant in most 
traditional texts. On the other hand, that ritual is the most generalized 
and important expression of the encompassment of minor gods by the 
major ones does not exclude other more direct expressions, through 
descent, names, or shared predicates (V. 109-110). Another way of 
relating inferior gods to superior ones is through the predicate of purity 
(ibid, and passim). As I make clear, none of these models are general- 
ized to include all gods (cf. V. 14); hence my case for encompassment 
rests more on what ritual shows than on the attempts at reflective con- 
ceptualization of Hawaiian priests and wise men. Indeed, I completely 
subscribe to Jane Harrison's view: "What a people does in relation to its 
gods must always be one clue, and perhaps the safest, to what it thinks" 
(Harrison 1903, vii). Furthermore, I have repeatedly mentioned that 
certain gods are outside the official hierarchy of the cults: the akua 
'unihipili (V. 30), the akua lele (V. 351 n. 31), and at least some of the 
sorcery gods (V. 42, 138, 370 n. 32). (No doubt Chariot's "wandering 
spirits" may also be included in the list.) I have also stressed that the 
goddesses "are not as hierarchized" as the gods (V. 113, cf. 19), a fact 
that has ritual correlates (V. 127). Indeed, I have not attempted to 
include the goddesses in my simplified summary of "the hierarchy of the 
gods and the hierarchy of men" (V. 109). But Chariot ignores all this 
and isolates a single sentence of that summary (the only sentence in my 

180 Pacific Studies, Vol. 10, No. 2— March 1987 

book to appear so extreme), disregarding, too, that I qualify it by saying 
that it is "a simplified model" and that "the reality is more complex" (V. 
110, cf. V.36). 

Let us now turn to the error (or rather errors) of fact made by 
Chariot. He claims that my statement, "there is no doubt that Ku, 
Lono, Kane, and Kanaloa are the highest gods" (V. 109), derives from 
"the conventional idea of 'the four great gods of Polynesia' " and espe- 
cially from Kamakau. First, I nowhere use the expression "the four 
great gods of Polynesia," which he gives as a quotation. Second, my 
statement derives neither from that "conventional idea" nor from 
Kamakau, but from the best available sources. One of the earliest 
Hawaiian sources, Malo himself, writes: "The names of the male deities 
worshipped by the Hawaiians, whether chiefs or common people, were 
Ku, Lono, Kane, and Kanaloa; and the various gods worshipped by the 
people and the ali'i were named after them" (V. 81). Note, incidentally, 
that the last sentence supports the view that minor gods were encom- 
passed by the four major ones, an additional proof that Chariot is 
wrong in attributing this theory to "systematizing nineteenth-century 
sources." More importantly, Malo's statement is fully supported by 
many prayers uttered during the luakini ritual that invoke the four 
main gods together (V. 269-270, 281-283, 290, etc.). These prayers are 
given, again, by the oldest sources: Kelou Kamakau and Malo. Those 
given by 1*1, who witnessed the rituals, also confirm that the four gods 
were worshipped together in the luakini temple. This evidence under- 
mines Chariot's curious thesis that "the association of four gods" is "an 
idea based on the Christian Trinity" (C. 125). Apart from all other dif- 
ferences with Christian Trinitarianism, I fail to see how a quaternity 
can be based on a trinity. Moreover, the fact that various quaternities 
are symbolically important in all Polynesian cultures is well known. 
Note also that Chariot attaches a different translation to the expression 
ke koko'oha o ke akua ("the quaternity of the god") from the one that I 
give ("the association of four gods," V. 13). No doubt Chariot wanted to 
suggest a connection with the "Trinity." 

Chariot claims that my use of this expression is "faulty" because the 
expression "oh association of four of the god(s)" is followed, in a chant, 
by the line "oh association of five of the god(s)" (Fornander 1916-1920, 
4:605). Thus, according to Chariot, "the chant is not referring to a sin- 
gle, overall supreme group, but to a number of groups" (C. 142 n. 27). 
This argument is strange. First, I have never suggested that other 
groups are not referred to in the chant, simply that the expression "oh 
association of four of the god(s)" refers to Ku, Kane, Lono, and Kana- 

Book Review Forum 181 

loa. Second, Chariot forgets that, with the exception of the line he 
quotes, all other numerical groups in the chant are multiples of four: 
forty thousand, four hundred thousand. This shows that the chant 
reflects in part the usual formula, "the 40000 deities, the 400000 deities, 
the 4000 deities" (cf. V. 13), which expresses the postulated equivalence 
of the totality of the gods and the numerical index four — the one, pre- 
cisely, which defines the group Ku, Lono, Kane, Kanaloa. I don't know 
what the "group of five" gods stands for, although I have noted that five 
is connected by one source with the god Ku (V. 350 n. 15). It may thus 
stand for five forms (or, as I call them, "particularizations") of Ku. 

Chariot's next statement that I adopt "the Trinitarian notions of other 
nineteenth-century Hawaiian writers, who degrade Kanaloa to a sort of 
demon" (C. 125) is completely false. Actually, my statement that "the 
quadripartition of the gods is a superficial phenomenon that conceals a 
tripartition at a deeper level" (V. 18), from which Chariot derives his 
extravagant interpretation, is the logical implication of the traditional 
pairing of Kane and Kanaloa as two sides of the same whole — a pairing 
reflected by the Kumulipo, in which the two gods appear as twin 
brothers. Perhaps Chariot believes here, for once, that even the Kumu- 
lipo is inspired by "the Trinitarian notions" of nineteenth-century 

The statements on the preeminence of Ku, Lono, Kane, and Kanaloa 
contained in the texts of Malo and Kelou Kamakau should be sufficient 
to demonstrate that the preeminence was real and traditional. Yet 
Chariot does not simply ignore that crucial evidence: He also falsely 
accuses me of leaving out "a good deal of evidence" that would prove 
that those four gods were not the highest. The first evidence that he 
mentions is a chant about the goddess Pele in which "those gods are 
mercilessly subordinated to her" (C. 125). But this chant dates from the 
late 1890s and was recorded in this century by Pukui (Pukui and Korn 
1973, 52). It mentions a geographical name (Borabora) not known tra- 
ditionally. Its reference to Pele as "ruler of the Menehune" (Pukui and 
Korn 1973, 55) in the context of a migration from Tahiti also betrays the 
fact that it is recent, since only in late nineteenth-century theories does 
one find mention of Menehune as migrants from the Society Islands to 
Hawaii (Barrere 1969, 41; cf. 36). Note also that only Ku and Lono, not 
Kane and Kanaloa, are put in a position inferior to Pele in the chant. In 
sum, this late and nontraditional text cannot support Chariot's claim 
that the four major gods were "mercilessly" subordinated to Pele in 
ancient Hawaii. 

More importantly, Chariot confuses different genres. The Pele litera- 

182 Pacific Studies, Vol. 10, No. 2— March 1987 

ture, especially in its revised, nineteenth-century form, reflects an anti- 
hierarchical bias that is perfectly consonant with the dominant role that 
goddesses (whose antistructural role I have emphasized) 13 play in it. 
Furthermore, tales in which not only inferior gods, but even men, sub- 
ordinate important gods and even ridicule them exist in Hawaii as else- 
where in Polynesia. But, as I have shown in a study of some of these 
tales (Valeri 1981), such explicit reversals are limited either to contexts 
in which they were ritually permitted (such as the Makahiki festival, 
when the "Pele literature" was performed in the dances), or to playful 
narration. Chariot, as usual, lumps all texts together without attempt- 
ing to establish beforehand their signification by an analysis of the 
genres and contexts in which they appear. He thus violates one of the 
basic rules of source criticism. No doubt the abolition in 1819 of the rit- 
ual system on which the hierarchy of gods was based changed the value 
of these playful reversals. By eliminating their very contrast with the 
serious contexts of temple ritual, it created a condition for transforming 
them into permanently valid charters of status for certain groups (par- 
ticularly in Ka'u). Precisely because of this, Chariot's use of some Pele 
chant as evidence against the idea that Ku, Lono, Kane, and Kanaloa 
were the highest gods is unsound both from a historical and a sociologi- 
cal point of view. Indeed, it is sufficient to remember that Pele is often 
considered as the daughter of Kuwahailo (Beckwith 1940, 30) — that is, 
Ku as introducer of human sacrifice (the very basis of temple ritual) — to 
put Chariot's claims in their true place. As a daughter of Ku, Pele can 
hardly be considered to dominate him, since she is genealogically 
encompassed by him. 

Chariot's second piece of evidence, the myth of Lonoka'eho's defeat 
by Kamapua'a, does not prove his point any better than the first. Lono- 
ka'eho is considered an ali'i, not a god, in the principal text of the 
Kamapua'a legend (Fornander 1916-1920, 5:327). Even the Kahiolo 
text, the only one to which Chariot refers, does not say that he is a god. 
At any rate, I have found no reference anywhere to Lonoka'eho as a god 
who is worshipped. More importantly, the Kamapua'a legend belongs 
to a playful and comic genre in which the desire for a reversal of 
hierarchical relations is given an imaginary satisfaction. Even if Lono- 
ka'eho were in the class of Lono gods, then, the legend would not prove 
the existence of a serious and ritually implemented alternative to the 
view that Lono is one of the supreme gods. 

As a further piece of evidence against the supremacy of the four gods, 
Chariot refers to a published work of his that I cannot find in any 
library to which I have access. I cannot therefore evaluate the examples 
he gives there. 

Book Review Forum 183 

Chariot also claims that since "it is impossible to absorb the female 
gods of Hawai'i into the four male gods" (C. 125), my view that inferior 
gods are encompassed by major ones is further invalidated. This is a 
strange argument, since I have explicitly said that goddesses form a class 
apart from the male gods and are themselves encompassed by Hina, 
who is paired with Ku. At the pinnacle of the Hawaiian pantheon, 
then, there is a male/female pair. Indeed I have written: "The structure 
of the pantheon — like that of the Kumulipo— reflects the primacy of the 
sexual principle" (V. 12). Hence Chariot's "deduction" that I "belittle" 
goddesses because I cannot reduce them to the major gods is totally 

Chariot argues against my statement that "goddesses are few and 
have a marginal position in the Hawaiian pantheon" (C. 125), saying: 
"Goddesses are in fact numerous and important." As I have already 
explained, Chariot misunderstands my use of the anthropological no- 
tion of marginality, which does not imply lack of importance, but an 
antistructural power. In noting the preponderant role that goddesses 
have in sorcery I have underscored that their importance consists pre- 
cisely in their power to transcend and threaten the official hierarchical 
system, not in the fact that they participate in it. Chariot, who naively 
equates "important" with "central," attempts instead to demonstrate 
that the goddess Pele and her priestesses "take part in the ritual" (of the 
luakini temple) . The "demonstration" is effected by referring to pages 
of my book that, as I have already noted, demonstrate nothing of the 
sort. Pele is not worshipped in the luakini temple proper and appears 
only at the conclusion of the ritual. Furthermore, I find no mention of 
"priestesses of Pele" in the texts to which I refer. Only po'e kaula ("seers" 
or "prophets," sex unspecified) are mentioned, but none of them offici- 
ate in the rite. Chariot's statements are a product of his fantasy, and I 
find it particularly objectionable that he refers to my book as evidence 
to support them. 14 

Chariot also criticizes my statement that Pele and the other goddesses 
are "ultimately controlled by the King" on the grounds that it contra- 
dicts "the Pele chants mentioned above in which supremacy is claimed 
for her" (C. 126). I don't see how chants that allegedly claim the 
supremacy of Pele over other gods can prove anything about who ulti- 
mately controls the cult of the goddess. Chariot confuses mythology 
with ritual. That Pele and other goddesses are ultimately controlled by 
the king is demonstrated by the fact that the worship of these goddesses 
is initiated at the beginning of the ritual year by a sacrifice consecrated 
by the king in an annex to his main temple (V. 328-329) . 

As for Chariot's statement that the goddesses are "numerous and 

184 Pacific Studies, Vol. 10, No. 2— March 1987 

important" — a statement for which he gives no evidence — I wonder 
what exactly he means. Given the immense number of Hawaiian 
deities, it is obvious that, in an absolute sense, goddesses are numerous. 
But when I say that "goddesses are few ... in the Hawaiian pantheon" 
(V. 19), I use "few" in a comparative sense. Indeed, the number of male 
deities, at least in our sources, is overwhelmingly superior to the num- 
ber of goddesses. As for the question of "importance," while, as I have 
said, some goddesses are important in an antistructural sense, a great 
many goddesses cannot be considered important in terms of Hawaiian 
values. This is because they are specifically defined as "patron deities" 
of certain groups of women involved in technical activities (such as 
beating and printing tapa cloth, see Malo 1951, 82) or even aesthetic 
ones (such as dancing, ibid.) that, however valued, do not have the 
same importance as the activities of war and production over which the 
great male gods preside. 15 

Finally, Chariot says that the idea that an individual's relation to the 
gods is mediated by the hierarchy is contradicted by numerous accounts 
of direct contact (C. 126). After reading this criticism, I began to doubt 
that Chariot had read my book in its entirety. Not only do I treat at 
length the kinds of direct contact to which he refers, and more, but I 
specifically state at different points that there are two rival modes of 
relating with the gods: one hierarchically mediated, the other direct (V. 
19-20, 138-140). 16 

The important point to grasp, however, is that direct contact through 
means such as dreams, visions, and descent is not in itself evidence of 
the absence of hierarchical encompassment. For instance, many of the 
family gods ('aumakua) result from the marriage of an important god 
(Chariot cites Ku) with a human woman (cf. NK, 1:36). The attributes 
of a local or familial god may be identical to some at least of the 
attributes of a more encompassing god and therefore provide a link 
between them. This link is often expressed in ritual subordination. 
Thus, as I have noted, the altars of local and familial fishing gods are 
reconsecrated yearly after the altar of the king s fishing god (a form of 
Ku) has been reconsecrated (V. 187, 378 n. 28). The same is true of the 
altars of local and familial agricultural gods, which can be reconsecrat- 
ed only after the king's agricultural temples (centered on Lono) have 
been rebuilt or reinaugurated (V. 187). The linkage between family 
cults and the cults of the society as a whole (controlled by the king) is 
explained by the fact that they both have analogous aims, which they 
realize at different social levels: they promote agriculture, fishing, and 
human fertility, sanction moral laws, and so on. As "families" are 

Book Review Forum 185 

"nested" in the social hierarchy, so family gods are "nested" in the gods 
of the global society. Direct relationships with certain deities thus imply, 
logically and usually ritually as well, indirect relationships with more 
distant and encompassing deities. 17 As the family cult is inconceivable 
without the cult of the society as a whole, so the family gods are incon- 
ceivable without the gods of the society as a whole. The relationship of 
family x 1 with the fishing deity y 1 is "direct," but at the same time impli- 
citly inscribed in its relationship with a larger social unit x 2 , which cor- 
responds to the fishing deity y 2 . 

In contrast to these relationships, many relationships with gods are 
only direct because they are individualistic or even antisocial (as in 
many cases of sorcery, V. 30, 33, 42, 138) or because they claim to tran- 
scend the social hierarchy (as in the case of many kdula, "seers," 
"prophets," V. 138-140). Chariot is unable to see the significance of 
these different types of relations with the gods because, among other 
things, his view of Hawaiian religion is completely asociological. He 
ignores the fact that representations are connected with actions and are 
actions themselves. Their meaning is therefore inseparable from the 
context of action, which includes the system of social relations. 

New peaks of misunderstanding are reached in Chariot's discussion of 
what he calls "the second major principle of Valeri's book," that is, my 
thesis that Hawaiian gods are essentially anthropomorphic. I leave out 
all the minor oddities 18 to concentrate on the basic point. Chariot makes 
an egregious error that completely vitiates his entire argumentation: He 
thinks that "anthropomorphic" simply means "having the physical form 
of man." He therefore claims that my point that Hawaiian gods are con- 
ceived anthropomorphically can only be proven if it can be proven 
"that all Hawaiian gods have human bodies" (C. 127). Since he himself 
believes that there are a "large number of gods that have only animal or 
elemental bodies" (ibid.), he concludes that my thesis is wrong. Chariot 
does not realize that, following anthropological and philosophical 
usage, I give "anthropomorphism" a much wider sense than the one 
that he gives to it. In Lalande's classic definition, for example, anthro- 
pomorphism "se dit de tout raisonnement ou de toute doctrine qui, pour 
expliquer ce qui n'est pas Fhomme (par exemple Dieu, les phenomenes 
physiques, la vie biologique, ia conduite des animaux, etc.) y applique 
de notions empruntees a la nature ou a la conduite humaine" (Lalande 

Another modern philosopher defines anthropomorphism as "that 
promiscuous mixing of our own intuitions of meaning, relevance, 
importance, with objective reality" (Taylor 1985, 1, 249). Thus if I 

186 Pacific Studies, Vol. 10, No. 2— March 1987 

interpret the behavior of an animal in terms of human feelings, inten- 
tions, reasoning, I anthropomorphize it. In other words, I give animal 
feelings and thoughts the form of human feelings and thoughts. Thus 
anthropomorphism does not necessarily imply the projection of the 
bodily form of man on something nonhuman, as Chariot seems to 
believe. 19 Throughout my book, I have made abundantly clear that the 
anthropomorphic character of Hawaiian gods is to be found, first and 
foremost, at the level of the signified, not that of the signifier: not in the 
gods' material manifestations, but in the human and social attributes 
they symbolize by means more complex than the purely iconic one 
offered by the human body. Thus I have repeatedly pointed out that 
most of these human attributes are signified by the natural, nonhuman 
"bodies" of the gods. 20 Indeed this is one of the central theses of the book 
that Chariot has managed to misunderstand. 

But Chariot is never happy with misunderstanding alone: he must 
also use objectionable means to achieve his aim. Thus he writes (C. 
126): "His [Valeri's] one argument [for anthropomorphism] is that all 
Hawaiian gods have a human body in their kino lau, their system of 
multiple bodies (9-12, 21, 31, 35, 47)," and he quotes my page 11: "the 
'genus' of all species included in one god belongs not to the natural 
world but to the human, social world" In this passage, Chariot gives 
the impression that the sentence he quotes from my book is equivalent 
to his own sentence, which precedes it. In other words, he is making me 
say that the genus of all species included in the god is the same thing as 
the human body that the god can assume. Chariot is confusing this 
statement, which is only due to him, with my thesis that the human 
body of a god is able to symbolize more clearly the genus (human in the 
cultural sense) that is also symbolized by the sum total of his natural 
bodies. At any rate the sentence that he quotes specifically refers not, as 
he says, to the human body of a god (who happens to be Kamapua'a), 
but to his porcine body. This shows, precisely, that the anthropomor- 
phic character of Kamapua'a does not depend on his being able to 
assume the physical body of man, but on the fact that his pig body "rep- 
resents human properties evoked by certain of the pig's qualities: viril- 
ity, activity, bellicosity, and so on" (V. 11). This is enough to dismiss as 
irrelevant Chariot's elucubrations on whether it is the pig body or the 
human body of Kamapua'a that is the principal one and similar argu- 
ments about other gods. Chariot is simply fighting figments of his own 
imagination, since my idea of what constitutes the "anthropomorphic" 
character of the gods is quite different from the one that he attributes 

Book Review Forum 187 

The same can be said of his statements on my supposed "presupposi- 
tion of a separation of human beings from 'nature,' " which, with char- 
acteristic illogic, he thinks is at the basis of my "strong anthropomorphi- 
zation of Hawaiian religion" (C. 127). As I have pointed out (V. 34), if 
there were such separation, there could be no anthropomorphization 
because natural phenomena could not signify human ones! Chariot 
attributes to me just the opposite of what I say. He then proceeds to defy 
decency when he defines the sentence where I have pointed out that no 
such separation exists as "the section in which Valeri admits that his the- 
ory cannot be found in the Hawaiian texts" (C. 128). The "theory" to 
which Chariot refers is Chariot's own theory of what my theory is. 21 

Let me repeat, then, once and for all, that my main thesis is that 
Hawaiian gods are anthropomorphic in the sense that their natural 
bodies are signs of human properties. But, as I just mentioned, I have 
also said that Hawaiian gods are usually anthropomorphic in another 
sense: they are able to manifest themselves in human form. However, I 
have made clear that the "human form" assumed by the gods must be 
conceived in its widest sense: not only as a human form spontaneously 
assumed when they appear in visions or dreams or even in physical pres- 
ence, but also as the human form that men give to the gods when they 
ritually incorporate them into anthropomorphic images or into human 
mediums (cf. V. 9 and my reference there to Firth 1930-1931; V. 72, 
345). Whatever the mode (spontaneous or contrived by man) and the 
quality (natural or artificial) of the gods' anthropomorphic manifesta- 
tion, it has the effect of making more evident the fact that all his non- 
human manifestations symbolize human predicates. In this sense, the 
frequently documented presence of a human body among the multiple 
bodies of gods is not without importance for my thesis, although it is not 
necessary to it. 

I readily admit that I cannot produce for all of the thousands of 
Hawaiian deities texts that state expressly that they have a human body 
as Chariot would like me to do. But such texts can be produced for a 
great many gods, and for all of the more important deities, male and 
female. 22 On the other hand, the few texts produced by Chariot as evi- 
dence of gods "with only animal bodies reported" either do not support 
or flatly contradict his point. 23 

Chariot asserts that "in accordance with his separation of human 
beings from nature and with his philosophical orientation, Valeri seeks 
to establish a nonnatural or 'supernatural,' invisible, immaterial realm 
or dimension" (C. 128). It is not clear how this supposed "dimension" 
could possibly be related to my supposed "separation of human beings 

188 Pacific Studies, Vol. 10, No. 2— March 1987 

from nature" or to my monistic philosophical orientation (made clear 
by my references to Feuerbach and Hegel). The important facts, how- 
ever, are as follows. As I have already indicated, I use the word "super- 
natural" only a couple of times and exclusively in quotation marks to 
indicate that I refer to another author's usage (as, precisely, in the sen- 
tence from V. 92 quoted on C. 128 as evidence), or that I am reluctant 
to use it. When I myself use it, I use it in its current anthropological 
sense, that is, as referring not to a nonnatural "dimension" of reality, 
but to a mode of operation that is different from the ordinary one and 
imputed to invisible divine action. It is in this sense that the term is used 
by anthropologically minded Polynesianists like Firth (1970). 

That the being of the gods is not reducible to that of their empirical 
manifestations is not an idea derived from my theory, contrary to what 
Chariot claims, but from the facts. Indeed, I do not think that anyone 
before Chariot (with the exception of the missionaries who accused 
Polynesians of idolatry or brute-worship) has ever doubted this. Malo, 
for one, writes: "All these gods, whether worshipped by the common 
people or by the ali'i, were thought to reside in the heavens. Neither 
commoner nor chief had ever discerned their nature; their coming and 
their going was unseen; their breadth, their length and their dimensions 
were unknown" (Malo 1951, 83). 

I do not think that these views have been influenced by Christian 
ideas; 24 on the contrary, comparative evidence from the Tikopia, a 
Polynesian people whose religion appears to have been uninfluenced by 
Christianity at the time they were studied by Firth, fully confirms its 
genuinely traditional character. Firth writes that "only in specific con- 
texts did the Tikopia attribute definite form to [the gods]. Linked with 
this view was the conception of atua as for the most part invisible to 
men. Hence the question of their 'proper' shape did not readily arise. 
Indeed, the absence of shape might be stressed — 'we do not see them; 
how do we know what they look like?' " (Firth 1970, 67; cf. 117-118). 
Moreover, Tikopia belief (cf. Firth 1967, 207) confirms my claim (and 
Malo's) that the gods "cannot be confused with those among their 
instantiations . . . that are supposed to empirically manifest the god's 
properties" (V. 32). 25 

Comparative evidence thus further belies Chariot's claim that Ha- 
waiians did not distinguish between the invisible reality of the gods and 
their visible manifestations (a thesis he needs in order to uphold his 
denial of the fundamentally anthropomorphic character of the gods). 
As I have myself pointed out, the two were strongly connected, particu- 
larly in ritual, but this is no justification for claiming that no distinction 
existed and therefore for maintaining that Hawaiian culture was char- 

Book Review Forum 189 

acterized by "its understanding of everything in physical terms" (C. 
136). Indeed, the latter claim shows to what an extent Chariot can pro- 
ject typically Western views (here the physicalist monism of modern sci- 
ence) onto Hawaiian ideas. 

Moreover, Chariot's thesis has the effect of attributing to Hawaiians 
his own illogic. Indeed, if the god were not something more than, say, 
an animal in which he manifests himself, then there would be no differ- 
ence between a mere animal and a god! Chariot mistakes the view that 
the god is only accessible through some empirical phenomenon for the 
view that the god only has empirical reality (cf. C. 129). By doing so he 
does serious injustice to the "considerable powers of abstraction" (Firth 
1970, 109) indicated by the Hawaiian, as by the Tikopia, notion 
of "god." 

Having established to his satisfaction that the Hawaiians were at his 
intellectual level, Chariot proceeds by attacking my use of the word 
"divine" in a section where I draw some preliminary conclusions. He 
claims that I introduce there a non-Hawaiian idea, because the word 
akua "is never . . . used as an abstraction, 'the divine' " (C. 129). But 
with one single (and partial) exception — to which I shall return and 
which is not the text quoted by Chariot — I have never used "the divine" 
as a translation of akua, only as a descriptive term that refers to an 
abstraction: the quality of "divinity" common to all gods, 26 that is, the 
quality that makes it possible to define them as "gods." 27 Although the 
existence of such a quality is clearly implied by the notion of akua, 
Chariot claims that it is a "non-Hawaiian idea" simply because it is not 
signified by a specific linguistic form. This is a very naive and incorrect 
view of the relationship between language and ideas. 

Not happy with formulating a disingenuous criticism alone, Chariot 
proceeds to misinterpret a sentence (V. 288) that refers to the transfor- 
mative relation existing betwen different gods or states of the same god 
in the luakini temple ritual. This transformative relation, I hypothesize, 
has the effect of suggesting to the audience a category of divine power 
more abstract than the individual gods. There is nothing particularly 
strange about this hypothesis. Anthropologists have noted that in poly- 
theistic religions one often finds, expressed in ritual form, extremely 
general ideas of divine power. This is the case, for instance, in the Chi- 
hamba ritual of the Ndembu (Turner 1962), or in the ida ritual of the 
Umeda of which it has been said: "By studying the sequence of ida as a 
whole one arrives rapidly at the idea that there are not many ritual fig- 
ures, but basically one such figure in process of transformation" (Gell 

Analogously, the sentence of my book quoted and criticized by 

190 Pacific Studies, Vol. 10, No. 2— March 1987 

Chariot (C. 130) refers to an implied category of divine power made 
apparent by ritual transformations, not, as he believes, to a cosmogonic 
argument in which "the gods emerge from an undifferentiated divine 
and merge back into it." Indeed, I refer there to a synchronic hierarchy 
of categories: In my view the category of divine power in the luakini rit- 
ual presupposes the different gods and is not separable from them. The 
idea of a "pulsation" back and forth from an "undifferentiated divine" 
cannot be found in my interpretation of the luakini ritual. On the con- 
trary, I make it clear that the different stages of the process of transfor- 
mation are always "frozen into personalities" (V. 288). Indeed, the 
transformation is summarized in my interpretation by the contrast 
between the gods Ku and Lono, not by the contrast between the undif- 
ferentiated and differentiated divine, which applies to the cosmogonic 

When I do claim that there is an analogy (but never an identity) 
between the ritual process and the cosmogonic process it is at a much 
higher level: that of the yearly ritual cycle. This is because the begin- 
ning of the year seems to be conceived as a return to the beginning of the 
world (V. 215). Since the god connected with the beginning of the year 
is Lono — whom a text (by K. Kamakau in V. 206) explicitly associates, 
in that particular temporal context, with a variety of attributes nor- 
mally associated with other gods; indeed, it associates it with the entire 
cosmos — I have hypothesized that the god represents at that moment 
the divine in general. I have also argued that, to some extent, the same 
is true of Ku in the luakini ritual, because this god often stands there for 
the other three major gods, whom it constantly implies, and for other 
reasons mentioned in my book. But the analogy between Lono or Ku 
and Po ("Night") as the most general metaphor of the divine is limited 
to what they refer to, since, as I have made abundantly clear, Po is not 
individualized, let alone personalized like Ku and Lono. Furthermore, 
none of these gods have the generality that the Po image has. I stand by 
this argument, against which Chariot offers no concrete criticism and 
which indeed he completely ignores. 

Having thus shown my actual views on the relationships between 
individual gods and a more generic category of divine power that has a 
variety of analogous (but by no means identical) ritual and cosmogonic 
expressions, let me turn to the basic point of contention between me and 
Chariot. This is my identification of the "undifferentiated divine" — 
that is, in my interpretation, the divine in its most generic form— with 
the Po ("Night") image. First let me note that Chariot is falsely giving 
the impression that my discussion of the age of Po has a great deal of 

Book Review Forum 191 

importance in my argumentation. This is not so. I say at the outset that 
I only give a "brief look at the genesis of the cosmos according to Hawai- 
ian mythology" (V. 3). Indeed, my discussion of the Kumulipo only 
takes up four pages at the beginning of my book. On pages 35-36 I ven- 
ture the hypothesis, which is presented as such, that Po is "the closest 
approximation to a supreme divine principle found in Hawaii" (V. 35), 
as it is perhaps in Maori cosmogony (Shortland 1882, 10). But whatever 
the merits of such a hypothesis, my interpretations of the Hawaiian 
religious system do not depend on it, contrary to what Chariot seeks 
to suggest by totally invalid or captious arguments, as I shall now 

Chariot's criticism of my association of Po with the "undifferentiated 
divine" is based on his usual error: he misconstrues a relative term as an 
absolute one. Although brief, my discussion of Po in the Kumulipo (V. 
7) shows that I view it as "the undifferentiated divine" only in a relative 
sense. It is "undifferentiated," first and foremost, relative to the myriad 
of individualized, personalized deities that appear in the age of Ao, 
"Light," which follows the age of Po, "Night," or "Darkness." Indeed, as 
I have pointed out (V. 30), it is believed that deities or ancestors who 
cease to be worshipped, or who leave the concrete form in which they 
can be approached, return to Po and dissolve in it (this "dissolution" of 
course is such only from the point of view of human perception, for 
which the world of Po is the "Unseen" [gloss of Handy and Pukui 1972, 
131]). This dissolution is clearly mentioned in the text to which I refer 
(V. 30) : "When the kahu or keeper felt it was unwise or even dangerous 
to keep the 'unihipili as a household presence, he could release the spirit 
and let it merge into the more tranquil eternity of Po" (NK, 1:196). 
Analogously, the spirits of the dead plunge from the leina (cliffs or sea- 
coast promontories) into Po, which Pukui defines as "measureless 
expanse of all space . . . timelessness of all time . . . eternity" (NK, 
1:35; cf. 40, 137). Po also stands for the generic divine in common 
expressions such as he ho'ike na ka po, "a revelation of the night," which 
Pukui explains as "a revelation from the gods in dreams, visions and 
omens" (Pukui 1983, 68). Since Po stands here for the gods in general, it 
can be called "undifferentiated [i.e., generic] divine" as I do in my 
book. In the expression mai ka po mai, translated as "from the gods, of 
divine origin" (PE, 307) or "out of the unseen" (Handy and Pukui 1972, 
131), Po has the same meaning of generic divine that it has in the pre- 
viously quoted expression. It indicates that something that is not indi- 
vidually identified belongs to the realm of the divine. Chariot neverthe- 
less claims that this expression does not support my view. 28 This is only 

192 Pacific Studies, Vol. 10, No. 2— March 1987 

because he understands neither the meaning of the expression nor my 

That I view the "undifferentiated divine" as undifferentiated only 
relative to its differentiation into individualized gods becomes perfectly 
clear once the passage of my book criticized by Chariot (C. 130) is put 
in its context: 

Until this point, the divine coincided first with the undifferen- 
tiated principle Po and then with its impersonal specifications 
presiding over the great divisions of nature. This identification 
of the undifferentiated divine with Po is made evident by the 
refrain that characterizes the age of Po: "the divine enters, man 
cannot enter. . . ." Being entirely divine, nature entirely ex- 
cludes man. By producing the first man, however, the divine 
brings about its own transformation. From now on it will be 
constituted by personal, anthropomorphic gods such as Kane 
and Kanaloa. Moreover, as I shall demonstrate, these personal 
gods regroup the natural species on the basis of a human 
"moral" logic that takes the place of or modifies the "natural" 
classificatory logic that the Kumulipo identifies with the state 
of the divine until man appears on the scene. (V. 7) 

Chariot's selective style of quotation covers not only the fact that Po is 
called "undifferentiated divine" relative to the personal gods of the age 
of Ao, but also that I do not consider it internally undifferentiated. 
Indeed, in the passage just quoted I refer to the fact that Po includes its 
impersonal specifications presiding over the great divisions of nature. 
Before that I give them in detail and show that they are sexually paired 
couples in which Po divides itself and which generate the biological cos- 
mos (V. 4-5). Chariot's objections to my view of Po betray his total mis- 
understanding of it, since they seem to imply that I identify Po only 
with its absolutely initial state, prior to its differentiation into the above 
mentioned paired forms. 

Thus he criticizes my use of the refrain o ke akua ke komo, 'a'oe komo 
kanaka— which I translate as "the god [or the divine] enters, man can- 
not enter" (V. 4)— to prove the "identification of the undifferentiated 
divine with Po," by the strange argument that it cannot be applied to Po 
because, says Chariot, "when first used (line 39), that line is twenty- five 
lines away from the last mention of po (line 14)" (C. 130). Chariot's 
argument is erroneous with regard both to the Kumulipo and to the text 
of my book. The reason is that the refrain is a refrain, which means that 
it returns several times in the section of the chant describing the age of 

Book Review Forum 193 

Po. Therefore it refers to that age as a whole, not to a single mention of 
Po in a single line. This is precisely why I use the line as evidence for the 
fact that "the Po period is . . . entirely divine" (V. 4; cf. 7, 216, 222), 
not simply, as Chariot seems to believe, for the fact that Po as men- 
tioned for the first time in line 14 of the Kumulipo is divine. 

In sum, when I speak of Po as "undifferentiated divine" I refer princi- 
pally to the age of Po as a whole, in contrast to the age of Ao when per- 
sonal gods are differentiated; only secondarily do I refer to Po before it 
differentiates itself internally in its "impersonal specifications presiding 
over the great divisions of nature" (V. 7). The basic point missed by 
Chariot is that the Kumulipo transforms a relationship of logical inclu- 
sion (which implies that Po is viewed as "the realm of the gods," "per- 
taining to or of the gods" [PE, 307]; in sum as a metaphor for the 
generic divine) into a genetic relation (which implies that Po is viewed 
as the undifferentiated origin of the individual gods who become dis- 
tinct in the age of Ao, "Light" [V. 6-7], and therefore vision, distinct 
knowledge). 29 Analogously, the relationship between Po and its paired 
sexual specifications (symbolizing the great realms of animal life) is 
represented both as one of logical inclusion (as is made clear by their 
names, which all consist of the morpheme Po plus a specifying suffix, V. 
4-5) and one of genetic differentiation. 

Po is thus both past as generative principle and present as the most 
encompassing category of the divine. This is precisely why I have writ- 
ten that Po is "the undifferentiated creative origin of the cosmos, which 
continues to exist in transcendence [i.e., in the "unseen," another mean- 
ing of Po] as its perennial source" (V. 35) . Saying, as Chariot does, that 
this is an idea similar to "the Thomist description of God as creator and 
sustainer of the universe" is betraying a total misunderstanding of my 
argument and a profound ignorance of Thomist philosophy, for which 
God (a perfect and intelligent substance endowed with free will) creates 
and sustains the universe providentially, that is, in view of an end that 
coincides with him (Aquinas, Summa Contra Gentiles III, 1). None of 
these characterizations are implied in my account of Po. 

Let me now turn to some criticism of particular points. Chariot 
attacks me for translating the already mentioned expression 'o ke akua 
ke komo, 'a'oe komo kanaka (quoted on V. 4, 7, 216, 222) as "the divine 
enters, man cannot enter." Strictly speaking, this criticism applies to 
only one case (V. 7). Chariot fails to mention that I also translate the 
expression as "the god enters, man cannot enter" (V. 216, 222). These 
two translations are considered equally possible in my first quotation of 
the line (V. 4), although the translation "the god" is given preeminence 

194 Pacific Studies, Vol. 10, No. 2— March 1987 

("the god [or the divine] enters, man cannot enter"). The reason I use 
"the divine" as an alternative to "the god" in this expression is as fol- 
lows: Ke akua cannot refer to an individual, personal god (such as, say, 
Kane) when it occurs in an expression qualifying the age of Po, from 
which personal gods seem to be absent. Indeed, they are said to appear 
only at the beginning of the age of Ao, together with man (V. 6-7). It 
seems to me, therefore, that the expression ke akua is used in Hawaiian 
exactly like in Greek 6 tizoc,, where it means "the god," but in two senses: 
in the sense of an individual, named god, and in the sense of the power 
common to all gods. "Les diverses puissances surnaturelles dont la col- 
lection forme la societe divine dans son ensemble peuvent elles-memes 
etre apprehendees sous la forme du singulier, 6 dzoc,, la puissance divine, 
le dieu, sans qu'il s'agisse pour autant de monotheisme" (Vernant 1974, 
2:87). What Vernant says of the Greek 6 #e6<; applies perfectly well to 
the Hawaiian case. The abstract idea of divine potency is referred to by 
an individualizing expression that means "the god." But "the god," in 
this usage, refers neither to an individually named god, nor to a 
supreme god; it is just the generic idea of divine power. This is why I 
find it legitimate to translate ke akua as "the divine" in one context. 

Chariot also objects: "Far from referring to a single, all-encompass- 
ing, undifferentiated principle, po is being constantly paired — with ao 
in the structure of the whole chant, with lipo in lines 7-8, and with la, 
'day,' in line 10" (C. 131). Let us look into these claims. 

1. It is not clear what Chariot means by saying that po is paired 
"with ao in the structure of the whole chant." My understanding, how- 
ever, is that the age of Po is paired with the age of Ao only in the sense 
that the former precedes the latter and indeed generates it (V. 6). This 
confirms that Po is conceived as the encompassing generative principle 
relative to Ao. If there is "pairing" here, it is certainly not one that 
belies the all-encompassing status of Po. 

2. "Pairing" is a vague term: there can be pairing of opposites (such 
as "male" and "female," "dark" and "clear") or pairing of equivalents. 
Lines 7-8 simply use the device of poetic parallelism to enumerate a 
series of equivalents of Po, which are in a relationship of redundancy to 
it, not of complementary opposition: 

From the source in the dark (lipo) was the earth formed 
from the source in the night (po) was darkness formed. 

(Johnson 1981, 3) 

3. The pairing of la, "day," with po in line 10 demonstrates precisely 
the opposite of what Chariot claims, since it neutralizes their opposi- 
tion: o ka lipo o ka la, o ka lipo o ka po, "darkness of day, darkness of 

Book Review Forum 195 

night." In other words, the day is as dark as the night; there is no con- 
trast of night and day— darkness reigns supreme. Indeed the next line 
states: po wale hoH, which Johnson (ibid.) translates "of night alone" 
and Beckwith (1951, 58) "nothing but night." Thus Po is the true 
encompassing principle at this point, contrary to what Chariot claims; 
the very evidence that he gives to disprove my point proves it. 

Let me consider now Chariot's argument for denying my view of Po 
as "creative origin of the cosmos" (V. 35) . The argument is that since 
"the word po appears first in line 5— after a description of the turning of 
the earth and sky and the sun being in shadow to illuminate the moon" 
(C. 131), Po cannot be considered the primal principle. Heaven and 
Earth, who are husband and wife, are this principle; hence — argues 
Chariot — "the mating of the earth and sky" is the "origin of the uni- 
verse" (C. 131). Chariot's argument is unconvincing for several reasons: 

1. That the word po ("darkness") is mentioned a few lines after 
Heaven and Earth is not in itself proof that darkness appears after the 
"mating of the earth and sky." On the contrary, the first lines imply that 
darkness is present from the beginning because "At the time when the 
earth became hot/ At the time when the heavens turned about" (Kumu- 
lipo, lines 1-2, Beckwith's trans., p. 58) the sun was darkened and the 
moon shone, as at night. Thus Po, darkness, is indicated as truly pri- 
mordial. 30 

2. The Kumulipo does not describe the "origin of the universe" as a 
whole, as Chariot implies, but only of the biological universe, the "liv- 
ing universe," as I call it (V. 9). The inorganic universe is taken for 

3. In contrast to the explicit mention of the mating of the paired 
forms of Po to produce the biological cosmos, there is no explicit men- 
tion in the Kumulipo that the life-forms derive from the mating of 
Heaven and Earth (which is itself only implicit). This couple, therefore, 
has an unclear status in the chant; it is more a generic image of 
generativity (cf . V. 215) than a true ancestral couple. 

In sum, in treating Po as the ultimate source of the cosmos, I am in 
agreement with what the Kumulipo (like many similar Polynesian cos- 
mogonies) states. But stressing that Po is the initial source does not in the 
least imply asserting that sex and procreation have no role in the genera- 
tion of the universe, since Po includes its sexually paired forms. Indeed, 
Chariot's statement — "Because Valeri is replacing this two-source origin 
[the mating of earth and sky] with a single-source one [Po], he cannot 
use sex and procreation. He must use 'creation' or 'production' " (C. 
131) — ranks as perhaps the falsest in his "critique." 

Although I describe the cosmogonic process in the Kumulipo as gene- 

196 Pacific Studies, Vol. 10, No. 2— March 1987 

alogical and sexual, and I quote approvingly Beckwith's view that this 
process is "actuated by desire, which is represented by the duality of sex 
generation" (1919, 300, V. 5), Chariot has the gall to retort to my sup- 
posedly "creationist" interpretation of the Kumulipo that "the Kumu- 
lipo is, however, a chant of the procreation, not the creation, of the uni- 
verse. There is nothing other than late, biblically influenced Hawaiian 
texts to compare with the extended creationistic systems of Samoa and 
the Society Islands" (C. 131-132). That the Kumulipo chant describes 
"the sexual production of the cosmos" (V. 89) is precisely what I have 
argued in the book; moreover, in a subsequent paper I have myself 
drawn the contrast between the Kumulipo and the later Hawaiian 
biblically influenced texts or the Central Polynesian cosmogonies (Valeri 
1986). Chariot does not see (or does not want to see) that our only real 
disagreement on what the Kumulipo says concerns the stage at which 
procreation first appears. I claim that this happens as soon as Po, 
"Night, Darkness," divides itself in female and male forms (V. 4); he 
claims that it is with the "mating of the earth and sky" (C. 131). Since 
the two events are practically contemporary, there is very little differ- 
ence between our positions. On the other hand, I strongly disagree with 
Chariot's reductionist extension of the "procreational" model of the 
Kumulipo to the relation between the personal gods and their manifes- 
tations and to the rituals that I analyze in my book. 

The evidence indicates that he is in error: the relationship of per- 
sonal, individualized gods with nature (more generally, with the "phe- 
nomenal" world) cannot be reduced simply to a "procreational," "gene- 
alogical" relationship of gods with the phenomenal world. Indeed, in 
Hawaii, as in every other Polynesian culture, the relationship of the 
gods with the world takes a great variety of forms (cf. Firth 1970, 98- 
99). It is precisely because I recognize this fact, not because I rule out 
procreation when it exists, that I use the term "production" or even 
"creation." 31 I do not contrast these terms to "procreation" (for instance 
in a passage quoted above I speak of "sexual production" V. 5), and they 
are not in a relationship of logical exclusion with it, contrary to what 
Chariot asserts without demonstration. They are simply more generic 
terms covering the totality of the relations between the gods and what 
they bring about, by sexual and asexual means. 

In my book, I have referred to many cases of nonprocreational pro- 
duction of species by gods, for instance by transformation of parts of 
their bodies (e.g., V. 359 n. 74). 32 Moreover, nonprocreational accounts 
are often used as alternatives to procreational ones. Malo, for instance, 
writes: "In the genealogy of Wakea it is said that Papa [ Wakea's wife] 

Book Review Forum 197 

gave birth to these islands. Another account has it that this group of 
islands were not begotten, but really made by the hands of Wakea him- 
self" (Malo 1951, 3). The "labor model" testified by this text seems to be 
used particularly to account for the god's way to bring about the growth 
of food plants. Thus the lands where "the best time to plant was during 
the winter rains" were called "the lands cultivated by Kanepua'a [a 
god]" (Kamakau 1976, 25). Some of the works of the gods to which the 
prayers refer are digging "the earth to soften and pulverize it," watering 
the plants, and shading them against excessive sun (ibid., 27-29). This 
work is the god's share in producing the fruits of the earth; no mention 
of the gods mating with goddesses to make the plants grow exists in 
these prayers (even if goddesses are mentioned in one — Kamakau 1976, 
30). Other sources confirm the absence of mating. For instance, Malo 
(1951, 206-207) says that any of the four major gods could be wor- 
shipped by the farmers to obtain crops, but he does not mention that 
any goddess was paired with them. This is not what we would expect if 
Chariot's "pan-procreational" thesis was correct. On the other hand, I 
have noted, although not enough for Chariot's taste, that the Makahiki 
ritual indicates that the god's action on plant growth has a sexual com- 
ponent (V. 214, 222, 224). Thus I do not deny this component; I simply 
claim that it is arbitrary to reduce to it all the forms taken by the god's 
productive action. Chariot forgets that not even the Kumulipo accounts 
for everything with a procreational-genealogical model. As I have 
already noted, all of inorganic nature is not so derived. 

More importantly, Chariot's attempt to reduce all facets of Hawaiian 
religious ideology to the Kumulipo model is flawed because his failure 
to consider the chant's purpose keeps him from correctly assessing the 
significance of its genealogical idiom. As I have shown elsewhere (Valeri 
1986), the Kumulipo must be viewed as an incantatory formula (cf. 
Beckwith 1951, 36, 38), whose purpose is to establish the absolute legiti- 
macy of the ali'i for whom it was composed. His legitimacy is made 
unassailable by "naturalizing" the historical process that brought him to 
his exalted position. Such "naturalization" is obtained by two conver- 
gent means: by reducing the process of succession to mere genealogy, 
that is, to mere procreation, without taking into account all properly 
political, action-based events; and by connecting human genealogies to 
a genealogy of natural species. Thus the complex process of human his- 
tory (partly documented in narratives) is reduced to genealogy, which is 
then projected onto the entire cosmos. This results in making the ali'i for 
whose birth the chant was composed into the outcome and the summary 
of the entire biological universe, that is, absolutely unquestionable 

198 Pacific Studies, Vol. 10, No. 2— March 1987 

(Valeri 1986). 33 Accordingly, the genealogical idiom by which the 
Kumulipo links the divine world to the phenomenal one should not be 
taken too literally. Indeed, one can apply to this chant what has recently 
been said of an Indonesian cosmogonic tradition also dominated by 
genealogy: "genealogical images sound more like figurative expressions 
for relations of world creation rather than literal genealogical truths" 
(Hefner 1985, 202). 

While Chariot's exclusive focusing on the procreative model may be 
explained by a peculiar obsession, no amount of obsession can justify his 
absurd claim that "Valeri has managed to depict the pinnacle of Hawai- 
ian religion as a masculine creator- god" (C. 132). The only "proof" 
advanced for this claim is my description of goddesses as "marginal." I 
have already shown that Chariot misunderstands this. But even this 
misunderstanding does not explain how he can put forward such a 
claim when I have explicitly stated: "The structure of the pantheon — 
like that of the Kumulipo — reflects the primacy of the sexual principle. 
The duality of the sexes is in effect divinized in the couple Ku (male)/ 
Hina (female)" (V. 12). This statement stands true for me even when I 
recognize, in agreement with the evidence, the hierarchical asymmetry 
of male and female in Hawaiian culture. That no contradiction is 
involved here is shown, for instance, by what a modern Hawaiian 
scholar has to say about the pervasiveness of both gender dualism and 
gender asymmetry in the Kumulipo: "The dichotomous style of bal- 
anced opposition of the opening chant of the Kumulipo is a brilliant 
reduction of the theme and metaphysics of dualism within a compressed 
poetic context. In philosophically reducing all organic and abstract 
form to dualistic categorization and opposition, however, the ancients 
were inevitably to grant greater respect to the masculine component of 
the universe and human life and to diminish the importance of the femi- 
nine" (Johnson 1981, 29). 

Ignoring all this, and the even stronger evidence of gender asymme- 
try provided by ritual, Chariot claims that I downgrade "Hawaiian 
goddesses and women, imposing on them an old-fashioned Western 
image" (C. 132). 34 He tries to deny the fact that Hawaiians considered 
women as ritually impure and excluded them from temple ritual, but he 
has no documentary basis to do so. His claim that in reiterating this 
well-known fact I appear "to argue against [my] sources, imposing a 
one-source picture upon the considerable evidence for a two-source, 
sexual ritual (e.g., 206, 217, 219-220, 282, 288; cf. 302-303)" (C. 132- 
133), simply displays what I am forced to call his considerable lack of 
honesty. None of the pages of my book that he mentions provide evi- 

Book Review Forum 199 

dence for his claim. On pages 206, 217, and 219, 1 refer to the Makahiki 
festival which, as a ritual transgression of the ordinary system of wor- 
ship, suspends the separation of the sexes associated with that system 
and has sexual components, as I have myself emphasized. Indeed, on 
page 206 I mention the fact that all temple sacrifices were taboo during 
this period. Chariot's other references presumably are to the dangling 
penis of the man-god Kahoali'i (V. 282) and the "dangling penis" of the 
temple image before it was covered with a loincloth (V. 288). I fail to 
see how this is evidence for a "sexual ritual" 35 ; as I have shown, it is not 
sexuality but nudity, as symbol of the divine in its "untamed" state, 
which is significant in those contexts. Here again, Chariot sees too much 
sex in Hawaiian symbolism, perhaps due to his own Western bias. 
Hawaiians, however, do not seem to have believed that sexual inter- 
course was possible with a dangling penis! Chariot's reference to pages 
302-303 presumably concerns my discussion of the symbolism of the 
Hawaiian house; again, this does not constitute evidence that the 
luakini ritual is a "sexual ritual." 36 

In his discussion on the alleged participation of "priestesses" in the 
purification rites preceding the entrance of men into the luakini temple, 
Chariot fails to mention that no Hawaiian source refers to it, and that it 
is only alluded to by Emerson (a source that Chariot disparages) in a 
footnote to Malo's text. At any rate, the rite occurs before the main rit- 
ual, from which women are notoriously excluded on pain of death. Fur- 
thermore, my comment on Emerson's dubious piece of information in 
no way reflects a "negative view of women." I am also at a loss to dis- 
cover what speculation on page 277 reflects this supposed negative 
view. Such veiled remarks are hardly acceptable in scholarly argument. 

Chariot's final criticism is that I use a "death-rebirth idea" (C. 133) 
borrowed from Frazer and Sahlins to interpret Hawaiian sacrificial rit- 
ual. He uses my discussion of the kali'i rite as an example. In fact, I do 
say that the kali'i is a symbolic death, but I do not say that it represents 
"rebirth." Chariot misunderstands my interpretation, which is based on 
my more general idea that ritual avoids the occurrence of what it repre- 
sents (here the death of the king) by producing it fictitiously. The rite 
does not have to emphasize this result by representing it as a "rebirth," 
especially here where the effect sought is the "taming" of the king's vio- 
lence. Indeed, in the kali'i "to strike the king" (one translation of kali'i) 
is immediately equated to "to make the king" (another translation of 
kali'i). Chariot suggests that simple "surrender" to the king may be 
involved in the rite, but his interpretation does not stand up to the evi- 
dence, which shows that the alleged "surrender" is in fact an attack in 

200 Pacific Studies, Vol. 10, No. 2— March 1987 

which the king can be killed (V. 211). Moreover, if it were a surrender, 
why would it be followed by a sham battle between the king's party and 
the opposing party? 37 

While there is no trace of the so-called "death-rebirth" model in my 
analysis of the kalii rite, I do use this model to interpret certain rites of 
the luakini temple. Although my interpretation is presented as conjec- 
tural, I find it justified for reasons that I have already stated in my book 
and which I will not repeat here. More generally, I will say that, in 
Hawaii as elsewhere (cf. Turner 1977), sacrifice employs an implicit 
death-rebirth idiom. For, as I have demonstrated at length with argu- 
ments Chariot does not counter, Hawaiian sacrifice is based on the prin- 
ciple of substitution. To kill a victim who stands for the sacrifier implies 
that the latter dies symbolically. But since this death is fictitious, and its 
only purpose is to transform the state of the sacrifier, its outcome can be 
(although is not necessarily, as I have mentioned) represented as a 
rebirth in a new state. 

The argument implicit in sacrifice (a rite that has no place in 
Chariot's romanticized view of Hawaiian religion) finds its way into 
verbal utterances, contrary to Chariot's claims. Thus many sacrificial 
prayers associate the death of the victim who stands for the sacrifier 
with obtaining ola, "life," for him. Consider, for instance, these lines, a 
motto for one of my chapters: 

A hiki a ola 

no nei make ia oe e Lono 

Life is obtained 

by this death by you, o Lono (V. 200) 

This text confirms what anybody who is not deaf can hear cried out by 
all sacrificial ritual: that life and death are dialectically connected, not 
radically separated, contrary to Chariot's opinion. 38 1 find it peculiar, to 
say the least, that he uses the reaction of a modern Hawaiian audience 
as evidence for this alleged separation in ancient Hawaii. What contem- 
porary Hawaiians say is no evidence for what their ancestors thought 
two hundred years ago. Furthermore, their suspect overreaction to the 
suggestion that the two senses of the word make ("desire" and "death") 
are connected cannot be taken, without further argument, as evidence 
that no such connection exists. It could legitimately be taken as evidence 
of the contrary, unless, of course, one decides that Hawaiians have no 
unconscious. But I, for one, do not subscribe to the paternalistic West- 
ern idea that Polynesians are Arcadian nature children, without dark 

Book Review Forum 201 

undersides. To say with Chariot that ancient Hawaiians, who used 
human sacrifice and other violent rituals to obtain ola, "life," saw "life 
as health and vigor and joyous sexuality, and death as the opposite" (C. 
135), that they did not "see death in life and life in death," is to confuse 
them disrespectfully with modern California hippies. That some mod- 
ern Hawaiian youths are closer to their California counterparts than to 
their ancestors, I would not deny, but to use their views as evidence on 
traditional values seems to me particularly unacceptable. 

I will now briefly discuss the content of Chariot's appendix, which 
concerns my criticism of his essay, "The Use of Akua for Living Chiefs." 
I have never implied that Chariot's thesis also referred to dead ali'i. It is 
not clear, for that matter, how this supposed misunderstanding of his 
position "misdirects Valeri's discussion," since the note where Chariot 
presumably illustrates his claim only refers to a discussion of the divini- 
zation of living ali'i (V. 145) and to our disagreement on the proper 
translation of the word akua in the sentence he akua na ali'i o Kona (V. 
370 n. 37). In neither case do I imply that Chariot denies the well- 
known fact that dead ali'i were divinized. 

It is not really necessary to spend much time on Chariot's rebuttal of 
my criticism of his thesis that the use of the term akua to refer to living 
ali'i is an innovation due to Kamehameha and "used in the post-Kame- 
hameha period." The reason is very simple. Chariot does not even 
answer my basic and decisive criticism: How could Hawaiians begin 
addressing ali'i as akua precisely "when traditional religious concepts 
were undergoing a crisis (cf. Choris 1822, 123; von Chamisso 1864, 
4:133-140) and the ali'i were losing their sanctity in the eyes of the peo- 
ple" (V. 145)? "Chariot attributes no motivation to this supposed inno- 
vation and no cause for its alleged success in the proto-missionary 
period" (ibid.) . In particular, how could the usage of addressing the ali'i 
as akua have spread precisely when the Hawaiian aristocracy had 
ceased to believe in the gods and had abolished their cult? Chariot 
should know that explanation of human action requires the reconstruc- 
tion of motives. But, as I have already noted, his way of writing history 
is most unhistorical: it consists in creating unmotivated sequences of 
events. Yet every historian knows that it is not sequencing in time, but 
motivation or "causation" that constitutes historical explanation. 

To these arguments I would now add one more. Chariot's thesis rests 
on a devaluation of the texts of Malo and Kamakau, where one finds 
explicit reference to the fact that high-ranking ali'i could be addressed 
or referred to as akua. Chariot argues that these texts have no documen- 

202 Pacific Studies, Vol. 10, No. 2— March 1987 

tary value because they "could easily have been influenced by the post- 
Kamehameha practice" (C. 137). But with regard to Malo, at least, this 
argument holds no water: Malo was born in 1795 and thus had every 
chance to become aware of such an important innovation as that of giv- 
ing the title akua to living ali'i. He would have mentioned this innova- 
tion by Kamehameha. The same argument is valid, a fortiori, for the 
compilers of the Mooolelo Hawaii (1838) . 

Finally, Chariot has nothing to say against another of my basic 
claims, that whatever one wants to say about the texts where ali'i are 
explicitly called akua, "one cannot hope to solve the question of the 
'divinity' of the ali'i by considering some texts independently of the 
global ideology that has produced them. Notions are not expressed only 
in words" (V. 145). It is precisely because of this that I rest my case that 
the highest ali'i were considered akua on the abundant evidence demon- 
strating that the attributes and prerogatives of the ali'i were similar or 
identical to those of the gods (V. 145-153). Although Chariot's muteness 
in the face of the above arguments makes a detailed response to his reit- 
erations unnecessary, I cannot leave unchallenged some of the erroneous 
statements or fallacies in which he indulges. 

He claims that I try to prove that ali'i "were called gods during their 
lifetimes" because they were descended from the gods. What I actually 
claim is that descent from the gods establishes that high-ranking ali'i 
have qualities considered divine (V. 144). To my argument that ali'i who 
were given the proper names of their gods must have been considered 
divine, Chariot retorts that one thing does not follow from the other, as 
demonstrated by the fact that "Hispanics . . . call sons Jesus" without 
implying that they are divine. The objection would be valid if the Span- 
ish usage were comparable to the Hawaiian one. But it is not: the name 
Jesus may be given to any Spaniard irrespective of rank and is therefore 
totally unmarked, but the names of an ali'i's gods could only be given to 
him, as far as I know (see sources quoted in V. 145). Furthermore, I do 
not claim that the usage of calling ali'i "by their god's proper name" (V. 
145) necessarily indicates that they also receive the common name akua; 
rather, I say: "In my opinion the custom of naming kings after their 
gods attests to the belief that the king is a manifestation of his gods and 
is therefore himself a god relative to all other men" (V. 145). As I have 
made abundantly clear, my discussion of whether or not Hawaiian ali'i 
were called akua is secondary in my eyes because it is only part of the 
wider discussion of whether or not they were considered "gods" in the 
sense that they had divine qualities not available to inferior men. 

This brings me to another false statement by Chariot, who writes of 
"Valeri's purpose of demonstrating an absolute, not a qualified, applica- 

Book Review Forum 203 

tion of the word akua to a living chief" (C. 139). Chariot displays here 
an insufferable disregard for what I actually say, which is as follows: 
"Generally speaking, the opposition akua/kanaka, 'god'/'man,' seems to 
be relative when applied to ali'i" (V. 143); "some kings, at least, are 
called akua, 'gods.' This is because no sharp distinction is made between 
the gods and their closest manifestations among humans. Indeed, it 
seems that the opposition akua/kanaka is a relative one and that certain 
men may be called the gods of others" (V. 144). It is precisely for not 
having understood the relative character of the appellation akua that I 
have taken Chariot to task in my book (V. 144)! 
Finally, two small points: 

1 . Chariot objects to my interpretation of line 734 of Haul ka lani, 
which I take from note 1 to the Fornander text. The issue is whether or 
not the word akua in that line is a veiled reference to the ali'i of Hilo. 
Chariot claims that "the line makes perfect sense when taken literally," 
that is, when translated "blinded are the eyes of the gods with salt" (C. 
139). I fail to see how this literal translation can make "perfect sense": 
Who has ever heard that the eyes of Hawaiian gods were blinded with 
salt? In contrast, vanquished ali'i were often blinded. I trust the inter- 
pretation contained in the footnote of Fornander's collection because 
that interpretation is due to a respectable Hawaiian source, "J. P. Kulu- 
waimaka, a famed chanter" (Fornander 1916-1920, 6:368). 

2. Chariot complains that I do not take account of his objections to 
the chant of Kuali'i as a document of the traditional use of the term 
akua to refer to living ali'i. The reason for my neglect of his objections is 
that I do not find them convincing, particularly because he fails to give 
a motive to Kamakau's alleged interpolation of references to King 
Kuali'i as akua in the text of the chant (lines 593-594). 

In his conclusion, Chariot says that "Hawaiian religion can be seen as 
itself only if looked at closely and carefully, that is, following scholarly 
rules of interpretation and argument" (C. 137). This is the only state- 
ment of his with which I wholeheartedly agree. But I have shown that 
Chariot has rarely followed scholarly rules in his "review" of my book. 
His implicit suggestion that "Pacific studies" follow his own example 
would be its end as a serious intellectual enterprise. 


1. Another source that implies this and not simply the virginity of the female (as Chariot, 
n. 3, claims) is Fornander 1916-1920, 4:540. 

204 Pacific Studies, Vol. 10, No. 2— March 1987 

2. Incidentally, Chariot does not even refer to this source properly. He quotes it as Pukui, 
Haertig, and Lee 1978, 1:88-89. In fact, the passage is from the second, not the first vol- 
ume of the book, which was published in 1979, not in 1978. Nor was the first volume pub- 
lished in 1978; it appeared in 1972. 

3. However, Mary Pukui is quoted there as saying: "Hawaiians placed very high value on 
virginity when a girl was reserved for the ali'i. Ali'i were considered to be under the keep- 
ing of the gods" (NK, 1:201). Pukui seems here to establish a connection between a reli- 
gious fact (the ali'i are "under the keeping of the gods") and the requirement that their 
spouses be virgins, although later she speaks of virginity as a means of insuring legitimacy. 
The two views are not incompatible. 

4. Even a couple of examples quoted from another chapter are in fact repeated in the 
chapter where I criticize Chariot. 

5. Some Hawaiian words (such as 'aumakua) have a special form in the plural. In confor- 
mity with common practice concerning the use of foreign words in English sentences, here 
as in my book, I have treated those words as invariable and therefore used them in their 
singular form only. In the same vein, English and Americans always write "twenty lira," 
using the singular form instead of the plural form of the Italian word. 

6. Chariot (C. 117) observes that in the passage that he quotes, as in others (e.g., V. 306, 
330), I put "is born" in quotation marks. He objects to this because "the Hawaiian equiva- 
lents do not appear in the Hawaiian text." Chariot seems to be ignorant of the fact that 
quotation marks (or inverted commas) may be used to indicate a nonliteral statement. 
Obviously, the god is not born in a literal, ordinary sense, and this is why I use the expres- 
sion with quotation marks. Analogously, when I say that 'Umi is " 'reborn' as a noble" (V. 
278), I imply that he is "reborn" in a metaphoric sense only. Chariot thinks instead, quite 
gratuitously, that the inverted commas imply that I claim to be quoting from my sources. 

7. Let me remark in passing that Chariot attempts to support his interpretation with a 
reference to Ke'aulumoku's chant, which is precisely the kind of evidence that, when I use 
it, he finds objectionable because it "represents a very personal, uncommon viewpoint" 
(C. 110). 

8. The rite is not mentioned in the third source (Wilkes 1845), but this is an extremely 
abbreviated (less than three pages) description of the luakini temple ritual. 

9. "Thus for the sacrifice to be efficacious, it is necessary for the victim to be at once iden- 
tified with and distinct from the sacrifier" (V. 48). 

10. Also a result of his inattentive reading is the statement that my final (and ironic!) sen- 
tence refers to the logic of the entire system. It simply refers to the relationship between 
material and ideal conditions of the system, the discussion of which takes less than a page 
in my book. 

11. This marginality is also indicated by the fact that the prophets leave before the king 
consecrates the offerings of chicken (in part at least contributed by them) and dogs to the 
goddesses (V. 329). Technically, then, they do not participate in the sacrifice proper. 

12. The presuppositional nature of the notion of encompassment implies that it can even 
take the form of "encompassment of the contrary," Dumont's very definition of hierarchy 

Book Review Forum 205 

(Dumont 1966). In other words, even gods in stark contrast w'th the major ones may be 
viewed as encompassed by them. 

13. The antistructural role of Pele and the goddesses associated with her is also manifested 
by their role as akua noho (gods of possession), which is emphasized by Malo (1951, 116). 

14. Chariot (C. 126) also seems to suggest that I claim that only the four main gods partic- 
ipate in the luakini temple ritual; but since he himself refers to my book as evidence for the 
worship of other gods as well, he cannot be serious. 

15. I may be allowed, in this context, to react to another author's criticism. In an other- 
wise perceptive review (for which I am very thankful), Jocelyn Linnekin takes issue with 
my use of a quotation from Malo: "The majority of women . . . had no deity and just 
worshipped nothing" (Malo 1951, 82). Linnekin writes that this statement is inconsistent 
with the "long list of female deities" (Linnekin 1985, 789) that precedes it. But she disre- 
gards the fact that Malo says that most of the female deities he enumerates were wor- 
shipped by certain women only — in most cases women who were involved in specialized 
activities (medicine, sorcery, dancing, tapa-printing). Therefore his general statement 
does not contradict his list of female deities. 

16. It is ironic that Chariot accuse me of ignoring the direct relationship with "family 
gods," since my extensive analysis of these gods (V. 19-30) is preceded by a statement in 
which a "sharp contrast" is noted between the relation with the "great gods" and the rela- 
tion with the family gods ('aumakua). Of the former relation I say that it may or it may 
not be hierarchically mediated (V. 19), but that it always "presupposes the social totality, 
precisely because everybody may invoke them" (V. 19-20). In other words, the four main 
gods are gods of all Hawaiians and therefore index the maximal level of Hawaiian society. 
In contrast, the akua 'aumakua are gods of kinship groups (or even individuals) only, and 
are directly related to them. My treatment of akua 'unihipili and more generally of sorcery 
should leave no doubt about the importance I give to direct relations with the gods. The 
cases of dreams, visions, and marriage of gods with humans mentioned by Chariot are 
treated in my book (V. 20-21) . 

17. As is noted by Pukui et al., an 'aumakua can be " 'a spiritual go between,' passing on 
prayers to the akua" (NK, 1:35). Thus they mention "praying to the aumakua as link to the 
akua" (36), another indication of the encompassment of 'aumakua by the akua, often 
represented by descent (ibid.). 

18. I found the following statement rather entertaining: "Valeri's exclusive equation of 
'subject' with 'human' is unusual. Some worldviews recognize nonhuman subjects, such as 
angels and leprechauns" (C. 126). I naively thought that angels and leprechauns were 
imaginary creatures in which the human subject projected himself! 

19. My view of anthropomorphism is also the view that Raymond Firth found adequate to 
Tikopia religion: "Atua generally seem to have been thought of by the Tikopia as anthro- 
pomorphic in the sense that they were endowed with human characteristics in most con- 
texts of discussion" (Firth 1970, 67); yet "when they wished to manifest themselves they 
might assume alternative forms: they might inhabit an inanimate object ... or they 
might enter an animate body, as a bird or a human being" (V. 109). In other words, the 
anthropomorphic character is present even when the gods manifest themselves in natural 
objects (jakatino, a cognate of the Hawaiian kino): "It is not held that the object reveals 

206 Pacific Studies, Vol. 10, No. 2— March 1987 

the actual shape of the god; he is spoken of and treated as if he were anthropomorphic" 
(Firth 1967, 207). 

20. See for instance this sentence: "the kino lau of gods are constituted by the projection of 
human predicates and their subjects (individual or collective) on the species and other phe- 
nomena of the natural world that evoke them" (V. 1 1) . 

21 . Analogously, the contradiction that Chariot attributes to me in his note 32 is of his own 

22. I believe, moreover, that most 'aumakua gods could manifest themselves in human 
mediums, thereby assuming human shape. As ancestors, furthermore, they must have 
been able to appear in human form in dreams and visions. 

23. Let me consider these texts in the order in which they are given by Chariot: 

1. "Fornander 1918-1919, 366 (shark)." This text mentions a king (ali'i) of the sharks, 
not a god. His brother is said to be "a famous shark deity," not by the text but by its editor, 
Thrum. There is no reference anywhere that this supposed shark deity was not able to 
assume the physical form of man, contrary to what Chariot claims. 

2. "Green 1923, 16-17 (bird)." This text is simply an animal tale and does not say that 
the two birds it mentions are gods. 

3. "Green 1923, 44-45 (caterpillars)." Possibly Chariot's idea that the caterpillar men- 
tioned in this text is a god is based on the arbitrary analysis of the name of its species, 
kuawehi, into kua (=akua) and wehi. This analysis is an example of Chariot's category 
"too many meanings" and is therefore erroneous by his own standards. Pukui and Elbert 
do not analyze the word at all, but define it as "dark caterpillar resembling poko, 
cutworm" (PE, 158). 

4. "Green 1923, 46 (squid)." This text does not say the squid is a god. 

5. "Green 1926, 66-69 (rat and owl)." In this animal tale, Rat and Owl (capitalized in 
the text as proper names) are said to be kupua, "demigods" (PE, 171). The editor of the 
tale explains what this means: "the animals are represented as kupua, or beings who can 
take either animal or human form at will." This flatly contradicts Chariot's claim that the 
tale is evidence of gods who can only take animal form. Note, moreover, that the tale 
refers to Owl as "he kanaka mahia'i," "a man who farms," and to Rat as "he kanaka 
palaualeo,"" a lazy man." The use of kanaka, "human being" (PE, 118), leaves no doubt of 
the fact that these two kupua are conceived more anthropomorphically than theriomorph- 
ically, and that it is dangerous to infer animal character from an animal proper name. 

6. "Green and Pukui 1936, 174-175 (squid); 176-177 (fish)." I don't have access to this 

With the possible exception of the last, none of the sources cited by Chariot as evidence 
prove his point. Moreover, they have little value as evidence, since they were collected in 
the twentieth century. But Chariot is not afraid of anachronism, since he also cites "many 
contemporary Hawaiian religious experiences" as proofs of his point. It is on this faulty or 
anachronistic evidence that he bases his theory that "anthropomorphism is a later element 
in Hawaiian religion that was applied secondarilv to the older theriomorphic gods" (C. 
n. 33). 

Chariot also mentions "a useful list of categories ... in Fornander 1919-1920, 6:52- 
55" (ibid.). It is not clear what he means by "categories." The text — a mere fragment- 
lists various natural phenomena and claims, quite erroneously, that they are worshipped 
as such. It does not say that these phenomena are manifestations of deities, let alone that 
they are their only manifestations. Chariot's cavalier use of evidence and quotation is 

Book Review Forum 207 

again very much apparent here and in his other claim that I neglect the "gods who emerge 
with animal species in the Kumulipo before the birth of anthropomorphic gods and 
human beings (e.g., Klwa*a, line 366)." Where does the Kumulipo say that the bird kiwa'a 
or any other such species is a god, I pray? And what exactly are these "gods who emerge 
with animal species"? 

24. Indeed, they are confirmed by a statement of Hewahewa, the last high priest, in con- 
versation with Judd: "In conversation with Hewahewa today, he said they always thought 
that God lived in heaven, that they made the idol and presented offerings hoping the spirit 
would descend and take possession of the idol and give answer to their enquiries as to the 
pono and the hewa. The old people said God had done so formerly" (J. P. Judd, Notes on 
his tour of Oahu, beginning 27 March 1834, Hawaiian Mission Children's Society Library, 
Judd Papers). (This reference was kindly provided by Marshall Sahlins.) 

25. Incidentally, I nowhere "admit" that my view "cannot be found in the Hawaiian 

Chariot's accusation that my language is "very irregular" when speaking of "invisible" 
because I apply this term not only to the gods, but also to the ali'i or to Kahiki, is in fact 
directed against the English language, where "invisible" means both "that cannot be seen; 
that by its nature is not an object of sight" and "not in sight; not to be seen at a particular 
place or time, or by a particular person" (OED). More importantly, I make clear that the 
nonvisibility of ali'i from commoners and of Kahiki from Hawaii is used as an experiential 
analogue of the invisibility of the gods that Kahiki and ali'i are ultimately meant to evoke. 
Thus I call Kahiki a spatial "metaphor" of the divine origins (V. 8), and I define the pros- 
tration taboo (kapu moe) of the ali'i as "a means of making these sacred beings invisible by 
acting not on their persons, but on their beholders" (V. 147). For other methods of creating 
experiential analogues of invisibility, see pages 148, 268-269, 300-301, 323-325. Note also 
that I speak of "relative invisibility. ... of high-ranking ali'i" (V. 147), not of absolute 

Even on this question, Chariot does not fail to offer us an amusing example of his ten- 
dency to contradict himself from one sentence to another. Just after having written that 
"Kahiki is called 'invisible' [by Valeri] apparently because it can't be seen from Hawai'i (8- 
9)," he continues: "Kahiki must be so treated [by Valeri] — must be placed in a transcen- 
dental dimension rather than be accepted as a distant land within this universe . . ." 

Note also that I nowhere say "that Hawaiians had the concept of immateriality" (C. 
129). The origin of this extraordinary statement may be in Chariot's misunderstanding of 
a general statement of mine about the use of perfume in ritual in general: "Note also that, 
like music, speech, or color, perfume has the property, precious from a ritual standpoint, 
of evoking immateriality in materiality, abstraction and generality in the concrete and 
individual (cf. Lewis 1980, 69)" (V. 268). In the analysis of the Hawaiian fact, which 
immediately follows this general statement, I do not use the term "immaterial" at all. On 
the contrary, I say that the gods "exist in experience thanks to a contrast between 'absence 
from sight' and 'presence in smell' " (V. 268-269) . 

26. See, for instance, V. 153, where the sentence "the predicates of the divine" means: the 
predicates that characterize all the gods. 

27. Note also that I use the expression "the divine" very seldom and not "widely" as 
Chariot claims. Moreover, two out of five references given by Chariot for this use are 
bogus. Thus "the divine" is not mentioned on page 88; on page 90, as the reference to 

208 Pacific Studies, Vol. 10, No. 2— March 1987 

Dumont makes clear, it is used as an abbreviation of the expression "divine sphere," which 
is contrasted to the expression "human sphere." 

28. Chariot also manages to distort what I say about this expression, by failing to mention 
that I quote it with Pukui's translation "out of the unseen." This translation evokes, pre- 
cisely, cognitive undifferentiation: what remains unseen cannot be differentiated. Only 
insofar as the gods have emerged from the "unseen" (Po) can they be identified and there- 
fore differentiated one from another. 

29. Note that the Kumulipo — exactly like some Maori cosmogonies (cf. Taylor 1855, 14- 
16) — is n ot simply a cosmogony. It is also a gnoseology; it accounts for the possibility of 
knowing the divine. 

30. I now consider erroneous my statement that the Earth's slime is the kumu "source" of 
Po (V. 4). This interpretation was suggested to me by Beckwith's (1951) translation of lines 
6-7 of the Kumulipo and by her rather confusing comments on pages 44-45, which made 
me think that the slime was produced by the union of Sky and Earth. The new translation 
by Johnson (1981), which unfortunately I read only after my book was in press, unequivo- 
cally shows that Po is its own source: 

From the source in the slime was the earth formed 

from the source in the dark was darkness formed 

from the source in the night (po) was night (po) formed. (Johnson 1981, 3) 

Thus Po is not generated sexually by the "marriage" of Sky and Earth, but generates itself, 
asexually, as the last line clearly states. 

31. Note, however, that in none of the pages quoted by Chariot (V. 156, 7, 75) to support 
his statement that I "must use 'creation' or 'production' " (C. 131) do I use the word "cre- 
ation." With a characteristic non sequitur Chariot, after having assimilated my use of the 
word "creation" to that of a certain "Western scholar," claims (n. 42): "Valeri's text is a 
good example of the power of distortion of such use. [In fact my use is different.] Valeri's 
emphasis on creation— rather than procreation — entails his elevation of 'sight and intelli- 
gence' to 'what is most human.' " I fail to see the logical connection between Chariot's two 
claims. While the intestinal regions were indeed involved in intellectual processes, the 
privileged connection between "sight and intelligence" is demonstrated by the word 'ike, 
which means: "To see, know, feel, greet, recognize, understand ... to receive revelations 
from the gods; knowledge, understanding, recognition, comprehension and hence learn- 
ing; sense, as hearing or sight; vision" (PE, 90). The superiority of seeing (and hearing, 
which is associated with it in 'ike) in humans is explicitly claimed by at least one Polyne- 
sian text and cannot therefore be dismissed as a "typically Western" view: "The eyes and 
ears of man govern the muscles and head. If the eyes sleep, the ears are closed also; but if 
the ears hear a voice or sound, the eyes open. They are thus the guardians of the body, and 
see or hear things nigh or distant bv which the bodv may be injured" (White 1887-1890, 

32. The transformation of bodily parts of a god and other modes of nonprocreational pro- 
duction found in Hawaii are also common elsewhere in Polynesia, for instance in Tikopia 
(Firth 1970, 87; 66). Note that in Tikopia the most generalized idiom to account for the 
production of food plants by the gods is not procreational but defecational, if one may say 
so. In effect, these plants are considered to be the gods' excrements (Firth 1970, 66; Firth 

Book Review Forum 209 

33. This shows, incidentally, that the Kumulipo is more anthropocentric and sociocentric 
than Chariot would have it. 

34. His evidence for this alleged imposition of an (unspecified) old-fashioned Western 
image is a string of quotations, given out of context, which he thinks he can refute by writ- 
ing: "In fact, all the activities mentioned above, except childbearing, were performed by 
men as well" (C. 132). Chariot's quotations from my book do not say otherwise, since they 
refer to women's "predominant role" in dancing, to their "privileged relationship with the 
female deities of sorcery," and to "properly feminine activities." None of these expressions 
implies that these activities are exclusively feminine. At any rate, it is not clear how 
Chariot's statement "proves" that my picture of Hawaiian women is based on Western 
views of women. Furthermore, I object to his constant mode of argumentation: if there is 
some similarity, however vague, between an account of Hawaiian views and Western 
views, then the description is false. Such argument is unacceptable because it denies a pri- 
ori, and without demonstration, that points of similarity may exist between Hawaiian 
views and Western views. 

35. How can the luakini temple ritual be considered a "two-source, sexual ritual" when 
any participant who is caught having sexual intercourse with a woman is put to death? 
Such prohibition indicates the explicitly nonsexual character of the ritual. 

36. I dismiss as dubious Kamakau's statement that the temple images to the left of the 
altar were female, because he is the only author to claim so and because his account of 
whatever concerns the luakini temple and its ritual is often untrustworthy (cf . V. 335-336; 
382 n. 32). Furthermore, the earliest source (Samwell 1967, 1177-1178) does not support 
the view that images on the left side of the altar were in any way contrasted to those on the 
right side; the iconography confirms this, since it does not show that the left-side images 
had female traits. To my claim that "all surviving images are anthropomorphic," Chariot 
objects that "a number of nonanthropomorphic, undeterminable, and unshaped stone 
gods can be seen at the Bishop Museum" (C. n. 36). But if these stones are shapeless and 
undeterminable, how can Chariot determine that they are images of gods? Simply because 
it is said so by curatorial tradition? At any rate, a former curator, Brigham, seems to have 
had a different opinion on this matter. He noted that he never saw any carved image of 
animals (let alone of theriomorphic gods), with the single exception of fish (Brigham 1902, 
92-94). The existence of these few fish images, however, does not prove that there were 
purely theriomorphic gods. Indeed, fish-gods are explicitly given both human and fish 
form in myth (V. 76-79). Shark-gods, in particular, seem to have often been given this 
double nature (Beckwith 1940, 129-130, 138-139, 140-143, etc.). With the exception of a 
couple of fish images in the Bishop Museum, then, my statement that "all surviving images 
are anthropomorphic" (V. 9) remains correct. Note, moreover, that my statement is fol- 
lowed by the qualification that "sometimes nonanthropomorphic components are in- 
cluded." It is implied that these components may have motivated Malo's claim that there 
were cases of theriomorphism. The essential point, however, is that anthropomorphic 
images played an absolutely dominant role in Hawaiian ritual, particularly in the temple 
ritual (cf. Brigham 1902, 93). This fact supports my view that ritual is fully efficacious 
only when the god is made present anthropomorphically. But I have never denied that the 
gods were made present indexically as Chariot claims with his Laka example (which I 
myself give, V. 396 n. 177). On the contrary, I have mentioned a variety of purely indexi- 
cal signs of the gods in ritual (V. 267-269, 270, 272, 281, 300, 308)— only I do not call 
them "images," since, precisely, they are not icons! 

210 Pacific Studies, Vol. 10, No. 2— March 1987 

37. Chariot again displays his literal-mindedness in his objection that disassembling the 
image of Lonomakua cannot imply that this god is killed. First, let me note that no act 
occurring in the context of ritual can be considered as a purely material, technological act. 
The ritual disassembling of a god's image is not the same thing as the dismantling of some- 
thing without symbolic signification: the image represents the god and even embodies 
him. Second, the dismantling cannot be evaluated independently of its syntagmatic con- 
text. Since it follows the king's violent termination of the Makahiki, the undoing of the 
image of the god who functions as Lord of Misrule appears as much more than a material 
undoing. Contrary to what Chariot suggests, this hypothesis is not in the least contradicted 
by the fact that Lono returns to Kahiki, since this return is effected by the god's neutraliza- 
tion through his symbolic death (again, without rebirth). 

38. Chariot admits that Hawaiians were "acquainted" with the "death-rebirth theory" 
(C. 134), but claims that they made "sparing use" of it. His pseudo-ecological hypothesis 
to explain this alleged sparingness ("lack of winter and spring, planting obviously living 
taro-tops rather than dead-looking seeds") cannot be taken seriously. A great number of 
cultures, even in the Pacific and Indonesia, employ the death-rebirth metaphor although 
they lack winter and spring and cultivate tuberous plants instead of seeds (why these 
should look dead is a mystery: rice seeds, for instance, are often conceived as alive in 
Southeast Asia). Chariot's other hypothesis — that "strong and consequent dualism" in 
Hawaii rules out the death-rebirth model because it excludes any dialectical connection of 
life and death — is also contradicted by comparative evidence. No wonder, since I cannot 
think of a deeper misunderstanding of the relationship between opposites in a dualistic sys- 
tem! It remains to be demonstrated, moreover, that Hawaiian thought is as dualistic as 
Chariot makes it. Dualism is certainly not the onlv mode of Hawaiian culture. 


Barrere, Dorothy 

1969 The Kumuhonua Legends: A Study of late 19th Century Hawaiian Stories of 
Creation and Origins. Pacific Anthropological Records No. 3. Honolulu: Bishop 
Museum Press. 

Beaglehole, John C. , ed. 

1967 The Journals of Captain James Cook in his Voyages of Discovery. Vol. 3, The 
Voyage of the Resolution and Discovery, 1776-1780. Parts 1 and 2. Cambridge: 
Cambridge University Press. 

Beckwith, Martha 

1940 Hawaiian Mythology. New Haven: Yale University Press. 

1951 The Kumulipo: A Hawaiian Creation Chant. Chicago: University of Chicago 

Beckwith, Martha, ed. and trans. 

1919 "The Hawaiian Romance of Laieikawai" (from the Hawaiian text of S. N. 
Haleole, 1863). Bureau of American Ethnology, 33rd Annual Report, 1911- 
1912, 285-666. 

Best, Elsdon 

1924 Maori Religion and Mythology. Section 1. Dominion Museum Bulletin No. 10, 

Book Review Forum 211 

Brigham, William T. 

1902 "Stone Implements and Stone Work of the Ancient Hawaiians." Memoirs of the 
Bernice Pauahi Bishop Museum. Honolulu: Bishop Museum Press. 

Choris, Louis 

1822 Voyage pittoresque autour du monde, avec des portraits de sauvages d'Amerique, 
d'Asie, dAfrique et des ties du Grand Ocean; des paysages, des vues maritimes, 
et plusieurs objets d'histoire naturelle. Paris: Firmin Didot. 

Chamisso de Boncourt, L. K. Adelbert von 

1864 Chamissos Werke, vols. 3 and 4. Berlin: Deutsches Verlagshaus Bong. 

Douglas, Mary 

1966 Purity and Danger: An Analysis oj Concepts oj Pollution and Taboo. London: 
Routledge and Kegan Paul. 

Dumont, Louis 

1966 Homo Hierarchicus: Essai sur le systeme des castes. Paris: Gallimard. 

Dumezil, Georges 

1948 Mitra-Varuna: Essai sur Deux Bepresentations Indo-Europeennes de la Souve- 
rainete. Paris: Gallimard. 

Elbert, Samuel H. 

1956- "The Chief in Hawaiian Mythology." Journal oj American Folklore 69:99-113, 
1957 341-355; 70: 264-276, 306-322. 

Finley, M.I. 

1977 Aspects of Antiquity: Discoveries and Controversies. Middlesex: Penguin Books. 

Firth, Raymond 

1930- Totemism in Polynesia." Oceania 1:291-321, 377-398. 

1967 The Work of the Gods in Tikopia. 2d ed. London: Athlone Press. 
1970 Bank and Beligion in Tikopia. London: George Allen and Unwin. 

Fornander, Abraham 

1878- An Account of the Polynesian Bace. 3 vols. London: Triibner. 

1916- Fornander Collection of Hawaiian Antiquities and Folk-lore. Edited by T. A. 
1920 Thrum, translated by John Wise. Memoirs of the Bernice Pauahi Bishop 
Museum, 4-6. 

Girard, Rene 

1972 La Violence et le sacre. Paris: Bernard Grasset. 

Gell, Alfred. 

1975 Metamorphosis of the Cassowaries: Umeda Society, Language and Bitual. Lon- 
don: Athlone Press. 

Green, Laura 

1923 Hawaiian Stories and Wise Sayings. Collected and translated by L. S. Green, 

edited by Martha Warren Beckwith. Poughkeepsie: Vassar College. 
1926 Folk-tales from Hawaii. Poughkeepsie: Vassar College. 

Green, Laura S., and Mary Kawena Pukui 

1936 The Legend of Kawelo and Other Hawaiian Folk Tales. Honolulu: Territory of 

212 Pacific Studies, Vol. 10, No. 2— March 1987 

Handy, E. S. C, and Mary Kawena Pukui 

1972 The Polynesian Family System in Kau, Hawaii. Rutland, Vt., and Tokyo: 
Charles E. Turtle. 

Harrison, Jane 

1903 Prolegomena to the Study of Greek Religion. Cambridge: Cambridge University 

Humphreys, Christmas 

1985 Buddhism. Harmondsworth: Penguin Books. 

Hefner, Robert 

1985 Hindu Javanese: Tengger Tradition and Islam. Princeton: Princeton University 

Johansen, J. Prytz 

1954 The Maori and His Religion in Its Non- Ritualistic Aspects. Copenhagen: Munks- 

Johnson, Rubellite Kawena 

1981 Kumulipo: The Hawaiian Hymn of Creation. Honolulu: Topgallant. 

Kahiolo, G. W. 

1978 He moolelo no Kamapuaa: The story of Kamapuaa. Translated by Esther T 
Mookini and Erin C. Neizmen. Honolulu: Hawaiian Studies Program, Univer- 
sity of Hawaii, Manoa. 

Kamakau, Kelou 

1919- "No na oihana kahuna kahiko." In Fornander Collection of Hawaiian Antiquities 
1920 and Folklore, edited by T. A. Thrum, translated by John Wise. Memoirs of the 
Bernice Pauahi Bishop Museum 6:2-45. 

Kamakau, S. M. 

1961 The Ruling Chiefs of Hawaii. Honolulu: Kamehameha Schools. 

1964 Ka Po'e Kahiko: The People of Old. Translated by Mary Pukui, edited by 

Dorothy Barrere. Honolulu: Bishop Museum Press. 
1976 The Works of the People of Old: Na Hana a Ka Po'e Kahiko. Translated by Mary 

K. Pukui, edited by Dorothy Barrere. Bernice Pauahi Bishop Museum Special 

Publication No. 61. Honolulu: Bishop Museum Press. 

Kepelino, K. 

1932 Kepelino's Traditions of Hawaii. Edited by Martha W. Beckwith. Honolulu: 
Bishop Museum Press. 

Kirch, Patrick V. 

1985 Feathered Gods and Fishhooks: An Introduction to Hawaiian Archaeology and 
Prehistory. Honolulu: University of Hawaii Press. 

Lalande, Andre 

1956 Vocabulaire Technique et Critique de la Philosophic Paris: Presses Universitaires 
de France. 

Lewis, Gilbert 

1980 Day of Shining Red: An Essay on Understanding Ritual. Cambridge: Cambridge 
University Press. 

Book Review Forum 213 

Linnekin, Jocelyn 

1985 "Review of Kingship and Sacrifice: Ritual and Society in Ancient Hawaii" in 
American Ethnologist 12, (4): 788-790. 

Malo, David 

1951 Hawaiian Antiquities (Mooolelo Hawaii). Translated by N. B. Emerson. 2d ed. 
[1903] Bernice Pauahi Bishop Museum Special Publication 2. Honolulu: Bishop Mu- 
seum Press. 

NK. SeePukui, Haertig, and Lee. 

PE. See Pukui and Elbert. 

Pukui, Mary K. 

1983 'Olelo No'eau. Hawaiian Proverbs and Poetical Sayings. Bernice Pauahi Bishop 
Museum Special Publication No. 71. Honolulu: Bishop Museum Press. 

Pukui, Mary K., and Samuel H. Elbert 

1971 Hawaiian Dictionary. Honolulu: University Press of Hawaii. (Cited as PE.) 

Pukui, Mary K., E. W. Haertig, and Catherine A. Lee 

1972- Nana i ke kumu (Look to the source). Two volumes (cited as NK 1 and NK 2). 
1979 Honolulu: Queen Liliuokalani Children's Center Publications. 

Pukui, Mary K., and Alfons Korn, trans., eds. 

1973 The Echo oj Our Song: Chants and Poems of the Hawaiians. Honolulu: Univer- 
sity Press of Hawaii. 

Rank, Otto 

1925 "Der Doppelgaenger: eine psychoanalytische Studie." Leipzig, Internationales 
psychoanalytischer Verlag. 

Samwell, David 

1967 Journal. In The Voyage of the Resolution and the Discovery, 1776-1780. See 
Beaglehole 1967. 

Shortland, Edward 

1882 Maori Religion and Mythology. London: Longmans, Green and Co. 

Taylor, Richard 

1855 Te ika a Maui; or, New Zealand and Its Inhabitants: Origin, Manners, Customs, 
Mythology, Religion, Rites, Songs, Proverbs, Fables, and Language of the 
Maori. London: Macintosh. 

Taylor, Charles 

1985 Philosophical Papers 1: Human Agency and Language. Cambridge: Cambridge 
University Press. 

Thrum, Thomas C. 

1909 "Tales from the Temples, part III." Hawaiian Annual for 1909, 44-54. 

Tregear, Edward 

1891 The Maori-Polynesian Comparative Dictionary. Wellington: Lyon and Blair. 

Turner, Victor 

1962 "Chihamba the White Spirit." Rhodes-Livingstone Paper No. 33. 
1977 "Sacrifice as Quintessential Process: Prophylaxis or Abandonment?" History of 
Religions 16 (3): 189-215. 

214 Pacific Studies, Vol. 10, No. 2— March 1987 

Valeri, Valerio 

1981 "Pouvoir des dieux, rire des hommes: Divertissement theorique sur un fait 
hawaiien." Anthropologie et Societes 5(3): 11-34. 

1982 "The Transformation of a Transformation: A Structural Essay on an Aspect of 
Hawaiian History (1809-1819)." Social Analysis 10:3-41. 

1985 "The Conqueror Becomes King: A Political Analysis of the Hawaiian Legend of 
'Umi." In Transformations of Polynesian Culture, edited by Anthony Hooper and 
Judith Huntsman. Wellington: Polynesian Society. 

1986 "Chroniques Polynesiennes, or: Hawaiian Historical Ideas and the Practice of 
Diarchy." Paper prepared for the symposium on "Symbolism Through Time," 
Wenner-Gren Foundation for Anthropological Research, January 12-21, Fez, 

Vernant, Jean Pierre 

1974 Mythe et pensee chez les Grecs. Vol. 2. Paris: Petite Collection Maspero. 

White, John 

1887- Ancient History of the Maori. 6 vols. Wellington: Government Printer. 

Williamson, Robert W. 

1933 The Religious and Cosmic Beliefs of Central Polynesia. 2 vols. Cambridge: Cam- 
bridge University Press. 


John Chariot, Institute of Culture and Communication, East-West 
Center, 1777 East-West Road, Honolulu, Hawaii 96848, U.S. 

Katharine Luomala, Department of Anthropology, University of 
Hawaii at Manoa, Porteus Hall 346, 2424 Maile Way, Honolulu, 
Hawaii 96822, U.S. 

Keith L. Morton, Department of Anthropology, California State Uni- 
versity-Northridge, Northridge, California 91330, U.S. 

Doug Munro, School of Arts, Darling Downs Institute of Advanced 
Education, Toowoomba, Queensland 4350, Australia 

Valerio Valeri, Department of Anthropology, University of Chicago, 
1126 E. 59th Street, Chicago, Illinois 60637, U.S. 


Hawaiian Genealogies 

Two Volumes 

Edith McKinzie 

edited by Ishmael W. Stagner 

Hawaiian Genealogies is an invaluable resource for novice Hawaiian ge- 
nealogists. The three-volume series will contain genealogies published in 
Hawaiian newspapers up to 1949 when the last Hawaiian language paper 
ceased publication. The genealogies and their explanations are given in 
the original Hawaiian with English translations interspersed, thus retain- 
ing the uniqueness of the old language while adding present-day clarity. 
Many of the original printer's errors have been either corrected or foot- 

The first volume, covering the period from 1842 to 1896, contains all 
the royal lines plus many of the major chiefly lines. The lists have been 
cross-checked for accuracy using Kamakau, Fornander, and Malo wher- 
ever possible. A numbering system for the oral lists aids the checking 
and cross-referencing of names, and extensive footnotes explain some 
potentially confusing and inconsistent parts of the genealogical lists. The 
index of names is a great asset in family research. 

Volume two (1858-1920) contains some major priestly lineages un- 
available until now. Also included are two family genealogical chal- 
lenges, as they appeared in newspapers of the time. Both volumes con- 
tain lively insights into Hawaiian family life of earlier generations. 

Vol. 1, 128 pp., $12.95, paper. ISBN 0-939154-28-5 

Vol. 2, 168 pp., $15.95, paper. ISBN 0-939154-37-4 

Published by 

The Institute for Polynesian Studies 

Brigham Young University— Hawaii Campus 

Distributed by University of Hawaii Press, 
2840 Kolowalu Street, Honolulu, Hawaii 96822 

new from Si 

your guide 
to searching 


using a 



Our recently published 20-page 
booklet will show you how to turn your 
personal computer into an effective 
research tool. You can have access to 
more than 180,000 citations and abstracts 
from the worldwide literature of sociology 
as fast as you can type "symbolic inter- 
actionist" (or, if you prefer, "structural 

The Guide contains: 

■ Hints on search strategies 

■ A review of Boolean logic 

■ Sample searches conducted on 
user-friendly systems 

■ A glossary of terms, and more . . . 

To receive your free copy, simply 
return the coupon below. And, if you wish 
to attend an SA online user workshop, 
please be sure to mark the appropriate box. 
Mail to: sociological abstracts, inc. 
P.O. Box 22206 
San Diego, CA 92122-0206 

r- — — — 

Please send me one copy of 

your guide to m 

abstracts using a personal computer. 

Send to: 

Name-please print 

Street Address 



I wish to attend an online training session in my area □. 
The best time for me would be (give month) 


ISSN 0023-829 


your guide 
to searching 

Linguistics & Language 
Behavior Abstracts 

using a 


Our recently published 20-page 
booklet will show you how to turn your 
personal computer into an effective 
research tool. You can have access to 
more than 80,000 citations and abstracts 
from the worldwide literature of sociology 
as fast as you can type "sociolinguistics" 
(or, if you prefer, "discourse analysis"). 

The Guide contains: 

■ Hints on search strategies 

■ A review of Boolean logic 

■ Sample searches conducted on 
user-friendly systems 

■ A glossary of terms, and more . . 


Linguistics and Language Behavior 
Abstracts is published quarterly. Each 
issue is fully-indexed. 

□ 1 -year subscription, $147 

□ Separate yearly cumulative index, $53 
(if ordered with regular subscription, 
$48) (if ordered by non-subscriber, 

□ User's Manual, 2nd edition, $26.50 
(All cited prices include postage and 

To receive your free copy, simply 
return the coupon below. And, if you wish 
to attend an LLBA online user workshop, 
please be sure to mark the appropriate 

Please send me one copy of 

your guide to Linguistics & Language 

Behavior Abstracts using a personal 


Send to: 

Name — please print 

Street Address 

City State 

I wish to attend an online training session in my area □ 

■ Zip. 

The best time for me would be (give month) 


P.O. Box 22206, San Diego, CA 92122 

hiiV-i'V. 1 ' 

■ I